{"generatedAt":"2026-07-30","schemaVersion":1,"source":"https://allaiask.com/alternatives","license":"https://allaiask.com/terms","methodology":"Effort is a published rule (same provider = drop-in, fully OpenAI-compatible provider = config, partially compatible or reasoning-mode loss = code-change, different wire protocol/auth = rewrite), not a score. Closeness = 0.40*parity + 0.25*effort + 0.25*price advantage + 0.10*speed advantage, min-max normalised within each candidate set; a missing measurement is dropped and the remaining weights renormalised.","effortDistribution":{"config":32,"codeChange":6,"rewrite":0,"total":38},"models":[{"slug":"claude-fable-5","name":"Claude Fable 5","provider":"Anthropic","blendedPerM":20,"stayPut":false,"picks":{"closest":"claude-haiku-4-5","cheapest":"gpt-5-6-luna","fastest":"qwen3-8-30b","sameProvider":"claude-haiku-4-5"},"alternatives":[{"slug":"claude-haiku-4-5","name":"Claude Haiku 4.5","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-90,"speedDeltaPct":261,"contextDeltaTokens":-800000,"parityGaps":["Context drops from 1,000,000 to 200,000 tokens","Max output drops from 128,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.625,"closeness":0.7109421906250669,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-77.5,"speedDeltaPct":90.2,"contextDeltaTokens":-600000,"parityGaps":["Context drops from 1,000,000 to 400,000 tokens"],"parityWins":[],"parityScore":0.875,"closeness":0.7021282383865503,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-97.8,"speedDeltaPct":207.3,"contextDeltaTokens":-600000,"parityGaps":["Context drops from 1,000,000 to 400,000 tokens","Max output drops from 128,000 to 64,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.625,"closeness":0.6949108254145091,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-70,"speedDeltaPct":85.4,"contextDeltaTokens":-700000,"parityGaps":["Context drops from 1,000,000 to 300,000 tokens","Max output drops from 128,000 to 64,000 tokens"],"parityWins":[],"parityScore":0.75,"closeness":0.6685044036213248,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-97.2,"speedDeltaPct":295.1,"contextDeltaTokens":0,"parityGaps":["Max output drops from 128,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.75,"closeness":0.6437236366265772,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-83.1,"speedDeltaPct":178,"contextDeltaTokens":0,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":[],"parityScore":0.875,"closeness":0.6286691354662388,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-94,"speedDeltaPct":1582.9,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Anthropic does not"],"parityWins":[],"parityScore":0.375,"closeness":0.6013573931700651,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-77.5,"speedDeltaPct":34.1,"contextDeltaTokens":1000000,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":["Context grows from 1,000,000 to 2,000,000 tokens"],"parityScore":0.875,"closeness":0.6011722545411435,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","blendedPerM":11.25,"stayPut":false,"picks":{"closest":"gpt-5-6-terra","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"gpt-5-6-terra"},"alternatives":[{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-60,"speedDeltaPct":77.3,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.8458395135556946,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-96,"speedDeltaPct":186.4,"contextDeltaTokens":0,"parityGaps":["Max output drops from 128,000 to 64,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.7777777777777778,"closeness":0.8097047400505296,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-70,"speedDeltaPct":159.1,"contextDeltaTokens":600000,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":0.8888888888888888,"closeness":0.6669895278492456,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-95,"speedDeltaPct":268.2,"contextDeltaTokens":600000,"parityGaps":["Max output drops from 128,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":0.7777777777777778,"closeness":0.6597887644963649,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-60,"speedDeltaPct":25,"contextDeltaTokens":1600000,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":["Context grows from 400,000 to 2,000,000 tokens"],"parityScore":0.8888888888888888,"closeness":0.6504398179696977,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-89.3,"speedDeltaPct":1468.2,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.4444444444444444,"closeness":0.6403360143399084,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-96,"speedDeltaPct":5468.2,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","No vision input","Loses prompt caching","Loses the 50% batch discount","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.2222222222222222,"closeness":0.6339537507050197,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-98.8,"speedDeltaPct":2445.5,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","No vision input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.3333333333333333,"closeness":0.6271371732764682,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"Anthropic","blendedPerM":10,"stayPut":false,"picks":{"closest":"claude-sonnet-4-6","cheapest":"gpt-5-6-luna","fastest":"claude-haiku-4-5","sameProvider":"claude-sonnet-4-6"},"alternatives":[{"slug":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-40,"speedDeltaPct":31,"contextDeltaTokens":-200000,"parityGaps":["Context drops from 500,000 to 300,000 tokens"],"parityWins":[],"parityScore":0.875,"closeness":0.7769784023423861,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-95.5,"speedDeltaPct":117.2,"contextDeltaTokens":-100000,"parityGaps":["Context drops from 500,000 to 400,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.75,"closeness":0.7486375616955359,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-55,"speedDeltaPct":34.5,"contextDeltaTokens":-100000,"parityGaps":["Context drops from 500,000 to 400,000 tokens"],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":0.875,"closeness":0.7458690899104594,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-haiku-4-5","name":"Claude Haiku 4.5","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-80,"speedDeltaPct":155.2,"contextDeltaTokens":-300000,"parityGaps":["Context drops from 500,000 to 200,000 tokens","Max output drops from 64,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.625,"closeness":0.7301191748790938,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-66.3,"speedDeltaPct":96.6,"contextDeltaTokens":500000,"parityGaps":[],"parityWins":["Context grows from 500,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.7115317530093047,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-55,"speedDeltaPct":-5.2,"contextDeltaTokens":1500000,"parityGaps":[],"parityWins":["Context grows from 500,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.6949132465050566,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":12.5,"speedDeltaPct":-24.1,"contextDeltaTokens":-100000,"parityGaps":["Context drops from 500,000 to 400,000 tokens"],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":0.875,"closeness":0.6598293109694567,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-fable-5","name":"Claude Fable 5","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":100,"speedDeltaPct":-29.3,"contextDeltaTokens":500000,"parityGaps":[],"parityWins":["Context grows from 500,000 to 1,000,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.65,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","blendedPerM":4.5,"stayPut":false,"picks":{"closest":"gemini-3-6-flash","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"gemini-3-6-flash"},"alternatives":[{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-25,"speedDeltaPct":107.3,"contextDeltaTokens":-1000000,"parityGaps":["Context drops from 2,000,000 to 1,000,000 tokens"],"parityWins":[],"parityScore":0.9090909090909091,"closeness":0.825133338393365,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-87.5,"speedDeltaPct":194.5,"contextDeltaTokens":-1000000,"parityGaps":["Context drops from 2,000,000 to 1,000,000 tokens","Max output drops from 64,000 to 32,000 tokens","No audio input","No extended-thinking mode"],"parityWins":[],"parityScore":0.6363636363636364,"closeness":0.7533041772435384,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-90,"speedDeltaPct":129.1,"contextDeltaTokens":-1600000,"parityGaps":["Context drops from 2,000,000 to 400,000 tokens","No audio input","No extended-thinking mode","Loses the free tier"],"parityWins":[],"parityScore":0.6363636363636364,"closeness":0.7032218617730537,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":0,"speedDeltaPct":41.8,"contextDeltaTokens":-1600000,"parityGaps":["Context drops from 2,000,000 to 400,000 tokens","No audio input","Loses the free tier"],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":0.7272727272727273,"closeness":0.6868022797316684,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-90,"speedDeltaPct":4354.5,"contextDeltaTokens":-1868928,"parityGaps":["Context drops from 2,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input","No audio input","Loses prompt caching","Loses the 50% batch discount","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.2727272727272727,"closeness":0.6542376359532116,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-73.3,"speedDeltaPct":1154.5,"contextDeltaTokens":-1868928,"parityGaps":["Context drops from 2,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No audio input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.45454545454545453,"closeness":0.6444811687173853,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-97.1,"speedDeltaPct":1936.4,"contextDeltaTokens":-1868928,"parityGaps":["Context drops from 2,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input","No audio input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.36363636363636365,"closeness":0.6394002817687998,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-120b","name":"GPT-OSS 120B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-94.2,"speedDeltaPct":1318.2,"contextDeltaTokens":-1868928,"parityGaps":["Context drops from 2,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input","No audio input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.36363636363636365,"closeness":0.6236495576688997,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"grok-4-3","name":"Grok 4.3","provider":"xAI","blendedPerM":1.5625,"stayPut":false,"picks":{"closest":"grok-4-20-0309-reasoning","cheapest":"cerebras-gpt-oss-120b","fastest":"cerebras-gpt-oss-120b","sameProvider":"grok-4-20-0309-reasoning"},"alternatives":[{"slug":"grok-4-20-0309-reasoning","name":"Grok-4.20 Reasoning","provider":"xAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":92,"speedDeltaPct":-46.9,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.8636091256481515,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":37.6,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":["No vision input"],"parityWins":[],"parityScore":0.8,"closeness":0.8264521240933748,"scoredOn":["parity","effort","price"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":116,"speedDeltaPct":16.3,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.7114782866734208,"scoredOn":["parity","effort","price","speed"]},{"slug":"grok-4-20-0309-non-reasoning","name":"Grok-4.20","provider":"xAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":92,"speedDeltaPct":6.1,"contextDeltaTokens":0,"parityGaps":["Max output drops from 64,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.6,"closeness":0.7057651747897999,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-71.2,"speedDeltaPct":2400,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input"],"parityWins":[],"parityScore":0.4,"closeness":0.7051218556539456,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-23.2,"speedDeltaPct":604.1,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens"],"parityWins":[],"parityScore":0.6,"closeness":0.7026606821386231,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":188,"speedDeltaPct":-43.9,"contextDeltaTokens":1000000,"parityGaps":[],"parityWins":["Context grows from 1,000,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.6949238626627431,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-glm-4-7","name":"GLM 4.7 (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":52,"speedDeltaPct":1920.4,"contextDeltaTokens":-800000,"parityGaps":["Context drops from 1,000,000 to 200,000 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input"],"parityWins":[],"parityScore":0.4,"closeness":0.6614756054353148,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"deepseek-v4-pro","name":"DeepSeek V4 Pro","provider":"DeepSeek","blendedPerM":0.54375,"stayPut":false,"picks":{"closest":"glm-5-2","cheapest":"cerebras-gpt-oss-120b","fastest":"cerebras-gpt-oss-120b","sameProvider":null},"alternatives":[{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":295.4,"speedDeltaPct":null,"contextDeltaTokens":872000,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 1,000,000 tokens"],"parityScore":0.75,"closeness":0.8042367000887822,"scoredOn":["parity","effort","price"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":727.6,"speedDeltaPct":14.7,"contextDeltaTokens":272000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 400,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.795882318715797,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-17.2,"speedDeltaPct":85.3,"contextDeltaTokens":272000,"parityGaps":["No extended-thinking mode"],"parityWins":["Context grows from 128,000 to 400,000 tokens"],"parityScore":0.75,"closeness":0.7486537921134772,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-17.2,"speedDeltaPct":3502.9,"contextDeltaTokens":3072,"parityGaps":["Max output drops from 64,000 to 32,768 tokens","Loses prompt caching"],"parityWins":["Context grows from 128,000 to 131,072 tokens"],"parityScore":0.5,"closeness":0.7451253102075323,"scoredOn":["parity","effort","price","speed"]},{"slug":"grok-4-3","name":"Grok 4.3","provider":"xAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":187.4,"speedDeltaPct":44.1,"contextDeltaTokens":872000,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 1,000,000 tokens"],"parityScore":0.75,"closeness":0.7335416921480622,"scoredOn":["parity","effort","price","speed"]},{"slug":"grok-4-20-0309-reasoning","name":"Grok-4.20 Reasoning","provider":"xAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":451.7,"speedDeltaPct":-23.5,"contextDeltaTokens":872000,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 1,000,000 tokens"],"parityScore":0.75,"closeness":0.7136141812711003,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":520.7,"speedDeltaPct":67.6,"contextDeltaTokens":872000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.7114814927907225,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":1969,"speedDeltaPct":-35.3,"contextDeltaTokens":272000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 400,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7098354027594403,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"Anthropic","blendedPerM":6,"stayPut":false,"picks":{"closest":"gpt-5-6-luna","cheapest":"gpt-5-6-luna","fastest":"claude-haiku-4-5","sameProvider":"claude-opus-4-8"},"alternatives":[{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-92.5,"speedDeltaPct":65.8,"contextDeltaTokens":100000,"parityGaps":["No extended-thinking mode"],"parityWins":["Context grows from 300,000 to 400,000 tokens"],"parityScore":0.875,"closeness":0.7986400316848216,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-25,"speedDeltaPct":2.6,"contextDeltaTokens":100000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 400,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7958637895211291,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":66.7,"speedDeltaPct":-23.7,"contextDeltaTokens":200000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 500,000 tokens"],"parityScore":1,"closeness":0.7760452186939459,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-haiku-4-5","name":"Claude Haiku 4.5","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-66.7,"speedDeltaPct":94.7,"contextDeltaTokens":-100000,"parityGaps":["Context drops from 300,000 to 200,000 tokens","Max output drops from 64,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.625,"closeness":0.7301415895997259,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-43.8,"speedDeltaPct":50,"contextDeltaTokens":700000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.7115029979269256,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":87.5,"speedDeltaPct":-42.1,"contextDeltaTokens":100000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 400,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7098162300481482,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-25,"speedDeltaPct":-27.6,"contextDeltaTokens":1700000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.6949110480232428,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-fable-5","name":"Claude Fable 5","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":233.3,"speedDeltaPct":-46.1,"contextDeltaTokens":700000,"parityGaps":[],"parityWins":["Context grows from 300,000 to 1,000,000 tokens","Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.65,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","blendedPerM":4.5,"stayPut":false,"picks":{"closest":"gpt-5-6-luna","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"gpt-5-6-luna"},"alternatives":[{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-90,"speedDeltaPct":61.5,"contextDeltaTokens":0,"parityGaps":["Max output drops from 128,000 to 64,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.7777777777777778,"closeness":0.8097839356291578,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":150,"speedDeltaPct":-43.6,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.7598295873322529,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-25,"speedDeltaPct":46.2,"contextDeltaTokens":600000,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":0.8888888888888888,"closeness":0.6670512623480009,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-87.5,"speedDeltaPct":107.7,"contextDeltaTokens":600000,"parityGaps":["Max output drops from 128,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":0.7777777777777778,"closeness":0.6598690206322992,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":0,"speedDeltaPct":-29.5,"contextDeltaTokens":1600000,"parityGaps":["Max output drops from 128,000 to 64,000 tokens"],"parityWins":["Context grows from 400,000 to 2,000,000 tokens"],"parityScore":0.8888888888888888,"closeness":0.6504918028340195,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-73.3,"speedDeltaPct":784.6,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.4444444444444444,"closeness":0.6404396410143613,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-90,"speedDeltaPct":3041,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","No vision input","Loses prompt caching","Loses the 50% batch discount","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.2222222222222222,"closeness":0.6340356157511914,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-97.1,"speedDeltaPct":1335.9,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 128,000 to 32,768 tokens","No vision input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — OpenAI does not"],"parityWins":[],"parityScore":0.3333333333333333,"closeness":0.627277014823497,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","blendedPerM":3.375,"stayPut":false,"picks":{"closest":"gemini-3-1-pro","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"gemini-3-1-pro"},"alternatives":[{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":33.3,"speedDeltaPct":-51.8,"contextDeltaTokens":1000000,"parityGaps":[],"parityWins":["Context grows from 1,000,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.8449324625245355,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-83.3,"speedDeltaPct":42.1,"contextDeltaTokens":0,"parityGaps":["Max output drops from 64,000 to 32,000 tokens","No audio input","No extended-thinking mode"],"parityWins":[],"parityScore":0.7272727272727273,"closeness":0.7896251398608424,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-86.7,"speedDeltaPct":10.5,"contextDeltaTokens":-600000,"parityGaps":["Context drops from 1,000,000 to 400,000 tokens","No audio input","No extended-thinking mode","Loses the free tier"],"parityWins":[],"parityScore":0.6363636363636364,"closeness":0.7032047973971056,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":33.3,"speedDeltaPct":-31.6,"contextDeltaTokens":-600000,"parityGaps":["Context drops from 1,000,000 to 400,000 tokens","No audio input","Loses the free tier"],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":0.7272727272727273,"closeness":0.6867974949413639,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-86.7,"speedDeltaPct":2049.1,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input","No audio input","Loses prompt caching","Loses the 50% batch discount","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.2727272727272727,"closeness":0.6542246260848157,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-64.4,"speedDeltaPct":505.3,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No audio input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.45454545454545453,"closeness":0.6444570017443176,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-96.1,"speedDeltaPct":882.5,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input","No audio input","Loses prompt caching","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.36363636363636365,"closeness":0.63935793159844,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-36.3,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":["No vision input","No audio input","Loses prompt caching","Loses the 50% batch discount","Loses documented data-residency options","Loses a published SLA","Trains on API data by default (or undocumented) — Google does not"],"parityWins":[],"parityScore":0.36363636363636365,"closeness":0.632512498033825,"scoredOn":["parity","effort","price"]}]},{"slug":"grok-4-20-0309-reasoning","name":"Grok-4.20 Reasoning","provider":"xAI","blendedPerM":3,"stayPut":false,"picks":{"closest":"grok-4-3","cheapest":"cerebras-gpt-oss-120b","fastest":"cerebras-gpt-oss-120b","sameProvider":"grok-4-3"},"alternatives":[{"slug":"grok-4-3","name":"Grok 4.3","provider":"xAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-47.9,"speedDeltaPct":88.5,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.8835248624688756,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-28.3,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":["No vision input"],"parityWins":[],"parityScore":0.8,"closeness":0.8264279624893436,"scoredOn":["parity","effort","price"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":12.5,"speedDeltaPct":119.2,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.7114705277843366,"scoredOn":["parity","effort","price","speed"]},{"slug":"grok-4-20-0309-non-reasoning","name":"Grok-4.20","provider":"xAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":0,"speedDeltaPct":100,"contextDeltaTokens":0,"parityGaps":["Max output drops from 64,000 to 32,000 tokens","No extended-thinking mode"],"parityWins":[],"parityScore":0.6,"closeness":0.7057574516830175,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-85,"speedDeltaPct":4611.5,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input"],"parityWins":[],"parityScore":0.4,"closeness":0.7051105761997895,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-60,"speedDeltaPct":1226.9,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens"],"parityWins":[],"parityScore":0.6,"closeness":0.7026489307898044,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":50,"speedDeltaPct":5.8,"contextDeltaTokens":1000000,"parityGaps":[],"parityWins":["Context grows from 1,000,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.6949186040603189,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-glm-4-7","name":"GLM 4.7 (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-20.8,"speedDeltaPct":3707.7,"contextDeltaTokens":-800000,"parityGaps":["Context drops from 1,000,000 to 200,000 tokens","Max output drops from 64,000 to 32,768 tokens","No vision input"],"parityWins":[],"parityScore":0.4,"closeness":0.6614552032798784,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"mistral-large","name":"Mistral Large 3","provider":"Mistral","blendedPerM":3,"stayPut":false,"picks":{"closest":"mistral-small","cheapest":"ministral-8b","fastest":"cerebras-gpt-oss-120b","sameProvider":"mistral-small"},"alternatives":[{"slug":"mistral-small","name":"Mistral Small 3.1","provider":"Mistral","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-91.3,"speedDeltaPct":98.4,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.9008022580315975,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-medium","name":"Mistral Medium 3","provider":"Mistral","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-73.3,"speedDeltaPct":50.8,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.89282697921189,"scoredOn":["parity","effort","price","speed"]},{"slug":"codestral","name":"Codestral","provider":"Mistral","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-85,"speedDeltaPct":93.4,"contextDeltaTokens":124928,"parityGaps":["No vision input"],"parityWins":["Context grows from 131,072 to 256,000 tokens"],"parityScore":0.875,"closeness":0.8483061601155193,"scoredOn":["parity","effort","price","speed"]},{"slug":"ministral-8b","name":"Ministral 8B","provider":"Mistral","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-95,"speedDeltaPct":159,"contextDeltaTokens":0,"parityGaps":["Max output drops from 32,768 to 16,384 tokens","No vision input"],"parityWins":[],"parityScore":0.75,"closeness":0.803728351264611,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-85,"speedDeltaPct":106.6,"contextDeltaTokens":268928,"parityGaps":["Loses the free tier"],"parityWins":["Context grows from 131,072 to 400,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":0.875,"closeness":0.7986404050258809,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-60,"speedDeltaPct":1031.1,"contextDeltaTokens":0,"parityGaps":["Loses a published SLA","Trains on API data by default (or undocumented) — Mistral does not"],"parityWins":[],"parityScore":0.75,"closeness":0.7626474714659233,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":50,"speedDeltaPct":27.9,"contextDeltaTokens":268928,"parityGaps":["Loses the free tier"],"parityWins":["Context grows from 131,072 to 400,000 tokens","Max output grows from 32,768 to 128,000 tokens"],"parityScore":0.875,"closeness":0.7458728107381986,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-85,"speedDeltaPct":3916.4,"contextDeltaTokens":0,"parityGaps":["No vision input","Loses the 50% batch discount","Loses a published SLA","Trains on API data by default (or undocumented) — Mistral does not"],"parityWins":[],"parityScore":0.5,"closeness":0.7451105761997894,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"qwen3-8-max","name":"Qwen 3.8 Max","provider":"Qwen","blendedPerM":2.8000000000000003,"stayPut":false,"picks":{"closest":"qwen3-7-max","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"qwen3-7-max"},"alternatives":[{"slug":"qwen3-7-max","name":"Qwen 3.7 Max","provider":"Qwen","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":0,"speedDeltaPct":4.3,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.8659985541788778,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-7-plus","name":"Qwen 3.7 Plus","provider":"Qwen","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-60.7,"speedDeltaPct":78.7,"contextDeltaTokens":0,"parityGaps":["No extended-thinking mode"],"parityWins":[],"parityScore":0.8333333333333334,"closeness":0.8220936341334759,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-83.9,"speedDeltaPct":5112.8,"contextDeltaTokens":-124928,"parityGaps":["Context drops from 256,000 to 131,072 tokens","No vision input"],"parityWins":[],"parityScore":0.6666666666666666,"closeness":0.8117867340729298,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-57.1,"speedDeltaPct":1368.1,"contextDeltaTokens":-124928,"parityGaps":["Context drops from 256,000 to 131,072 tokens"],"parityWins":[],"parityScore":0.8333333333333334,"closeness":0.7959858459874052,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-glm-4-7","name":"GLM 4.7 (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-15.2,"speedDeltaPct":4112.8,"contextDeltaTokens":-56000,"parityGaps":["Context drops from 256,000 to 200,000 tokens","No vision input"],"parityWins":[],"parityScore":0.6666666666666666,"closeness":0.7681580194992952,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-23.2,"speedDeltaPct":null,"contextDeltaTokens":744000,"parityGaps":["No vision input","Loses documented data-residency options"],"parityWins":["Context grows from 256,000 to 1,000,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":0.6666666666666666,"closeness":0.7671961387029494,"scoredOn":["parity","effort","price"]},{"slug":"grok-4-3","name":"Grok 4.3","provider":"xAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-44.2,"speedDeltaPct":108.5,"contextDeltaTokens":744000,"parityGaps":["Loses documented data-residency options"],"parityWins":["Context grows from 256,000 to 1,000,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":0.8333333333333334,"closeness":0.7668823037415464,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-95.3,"speedDeltaPct":2283,"contextDeltaTokens":-124928,"parityGaps":["Context drops from 256,000 to 131,072 tokens","No vision input"],"parityWins":[],"parityScore":0.6666666666666666,"closeness":0.7605798347020755,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","blendedPerM":2.15,"stayPut":false,"picks":{"closest":"grok-4-3","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":null},"alternatives":[{"slug":"grok-4-3","name":"Grok 4.3","provider":"xAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-27.3,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.923507775621211,"scoredOn":["parity","effort","price"]},{"slug":"grok-4-20-0309-reasoning","name":"Grok-4.20 Reasoning","provider":"xAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":39.5,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.9034996285912825,"scoredOn":["parity","effort","price"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":57,"speedDeltaPct":null,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.7871468621953849,"scoredOn":["parity","effort","price"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":109.3,"speedDeltaPct":null,"contextDeltaTokens":1000000,"parityGaps":[],"parityWins":["Context grows from 1,000,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.7714818009728512,"scoredOn":["parity","effort","price"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-93.9,"speedDeltaPct":null,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens"],"parityWins":[],"parityScore":0.5,"closeness":0.7212337957970911,"scoredOn":["parity","effort","price"]},{"slug":"gpt-oss-120b","name":"GPT-OSS 120B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-87.8,"speedDeltaPct":null,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens"],"parityWins":[],"parityScore":0.5,"closeness":0.719406704526394,"scoredOn":["parity","effort","price"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-79.1,"speedDeltaPct":null,"contextDeltaTokens":-868928,"parityGaps":["Context drops from 1,000,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens"],"parityWins":[],"parityScore":0.5,"closeness":0.7168008530419572,"scoredOn":["parity","effort","price"]},{"slug":"deepseek-v4-pro","name":"DeepSeek V4 Pro","provider":"DeepSeek","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-74.7,"speedDeltaPct":null,"contextDeltaTokens":-872000,"parityGaps":["Context drops from 1,000,000 to 128,000 tokens","Loses the free tier"],"parityWins":[],"parityScore":0.5,"closeness":0.7154829511417822,"scoredOn":["parity","effort","price"]}]},{"slug":"claude-haiku-4-5","name":"Claude Haiku 4.5","provider":"Anthropic","blendedPerM":2,"stayPut":false,"picks":{"closest":"gpt-5-6-luna","cheapest":"mistral-small","fastest":"gemini-3-5-flash-lite","sameProvider":"claude-sonnet-4-6"},"alternatives":[{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-77.5,"speedDeltaPct":-14.9,"contextDeltaTokens":200000,"parityGaps":[],"parityWins":["Context grows from 200,000 to 400,000 tokens","Max output grows from 32,000 to 64,000 tokens"],"parityScore":1,"closeness":0.8486613666092847,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-sonnet-4-6","name":"Claude Sonnet 4.6","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":200,"speedDeltaPct":-48.6,"contextDeltaTokens":100000,"parityGaps":[],"parityWins":["Context grows from 200,000 to 300,000 tokens","Max output grows from 32,000 to 64,000 tokens"],"parityScore":1,"closeness":0.8270002292478681,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":125,"speedDeltaPct":-47.3,"contextDeltaTokens":200000,"parityGaps":[],"parityWins":["Context grows from 200,000 to 400,000 tokens","Max output grows from 32,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7958884022930998,"scoredOn":["parity","effort","price","speed"]},{"slug":"claude-opus-4-8","name":"Claude Opus 4.8","provider":"Anthropic","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":400,"speedDeltaPct":-60.8,"contextDeltaTokens":300000,"parityGaps":[],"parityWins":["Context grows from 200,000 to 500,000 tokens","Max output grows from 32,000 to 64,000 tokens"],"parityScore":1,"closeness":0.7760952233856515,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":-71.9,"speedDeltaPct":9.5,"contextDeltaTokens":800000,"parityGaps":[],"parityWins":["Context grows from 200,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.748756060849517,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-small","name":"Mistral Small 3.1","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-86.9,"speedDeltaPct":-18.2,"contextDeltaTokens":-68928,"parityGaps":["Context drops from 200,000 to 131,072 tokens","Loses prompt caching"],"parityWins":["Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.7142857142857143,"closeness":0.736530219914632,"scoredOn":["parity","effort","price","speed"]},{"slug":"codestral","name":"Codestral","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-77.5,"speedDeltaPct":-20.3,"contextDeltaTokens":56000,"parityGaps":["No vision input","Loses prompt caching"],"parityWins":["Context grows from 200,000 to 256,000 tokens","Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.7142857142857143,"closeness":0.7340438958573913,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-medium","name":"Mistral Medium 3","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-60,"speedDeltaPct":-37.8,"contextDeltaTokens":-68928,"parityGaps":["Context drops from 200,000 to 131,072 tokens","Loses prompt caching"],"parityWins":["Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.7142857142857143,"closeness":0.7285801544869489,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","blendedPerM":0.45,"stayPut":false,"picks":{"closest":"gpt-5-6-terra","cheapest":"mistral-small","fastest":"gemini-3-5-flash-lite","sameProvider":"gpt-5-6-terra"},"alternatives":[{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":900,"speedDeltaPct":-38.1,"contextDeltaTokens":0,"parityGaps":[],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.8458818738284103,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-sol","name":"GPT-5.6 Sol","provider":"OpenAI","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":2400,"speedDeltaPct":-65.1,"contextDeltaTokens":0,"parityGaps":[],"parityWins":["Max output grows from 64,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7598348371521597,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":650,"speedDeltaPct":-9.5,"contextDeltaTokens":600000,"parityGaps":[],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.711483572785933,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-small","name":"Mistral Small 3.1","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-41.7,"speedDeltaPct":-4,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","Loses prompt caching"],"parityWins":[],"parityScore":0.625,"closeness":0.7007991855134915,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":25,"speedDeltaPct":28.6,"contextDeltaTokens":600000,"parityGaps":["Max output drops from 64,000 to 32,000 tokens"],"parityWins":["Context grows from 400,000 to 1,000,000 tokens"],"parityScore":0.875,"closeness":0.6987408668872027,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-1-pro","name":"Gemini 3.1 Pro","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":900,"speedDeltaPct":-56.3,"contextDeltaTokens":1600000,"parityGaps":[],"parityWins":["Context grows from 400,000 to 2,000,000 tokens"],"parityScore":1,"closeness":0.6949299411959505,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-medium","name":"Mistral Medium 3","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":77.8,"speedDeltaPct":-27,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","Loses prompt caching"],"parityWins":[],"parityScore":0.625,"closeness":0.6928536190208758,"scoredOn":["parity","effort","price","speed"]},{"slug":"mistral-large","name":"Mistral Large 3","provider":"Mistral","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":566.7,"speedDeltaPct":-51.6,"contextDeltaTokens":-268928,"parityGaps":["Context drops from 400,000 to 131,072 tokens","Max output drops from 64,000 to 32,768 tokens","Loses prompt caching"],"parityWins":[],"parityScore":0.625,"closeness":0.663981628939775,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"gpt-oss-120b","name":"GPT-OSS 120B","provider":"Groq","blendedPerM":0.26249999999999996,"stayPut":false,"picks":{"closest":"gpt-oss-20b","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"gpt-oss-20b"},"alternatives":[{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":-50,"speedDeltaPct":43.6,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.9439074827481133,"scoredOn":["parity","effort","price","speed"]},{"slug":"qwen3-8-30b","name":"Qwen 3.8 30B","provider":"Groq","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":357.1,"speedDeltaPct":-11.5,"contextDeltaTokens":0,"parityGaps":[],"parityWins":[],"parityScore":1,"closeness":0.9126651901017625,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":71.4,"speedDeltaPct":214.1,"contextDeltaTokens":0,"parityGaps":["Loses the 50% batch discount"],"parityWins":[],"parityScore":0.8333333333333334,"closeness":0.8784588648840791,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-glm-4-7","name":"GLM 4.7 (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":804.8,"speedDeltaPct":153.8,"contextDeltaTokens":68928,"parityGaps":["Loses the 50% batch discount"],"parityWins":["Context grows from 131,072 to 200,000 tokens"],"parityScore":0.8333333333333334,"closeness":0.8347930054132755,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":719,"speedDeltaPct":null,"contextDeltaTokens":868928,"parityGaps":["Loses the 50% batch discount","Loses documented data-residency options"],"parityWins":["Context grows from 131,072 to 1,000,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":0.6666666666666666,"closeness":0.767197230011717,"scoredOn":["parity","effort","price"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":1614.3,"speedDeltaPct":-90,"contextDeltaTokens":268928,"parityGaps":["Loses the free tier"],"parityWins":["Context grows from 131,072 to 400,000 tokens","Max output grows from 32,768 to 128,000 tokens"],"parityScore":0.8333333333333334,"closeness":0.729198864202619,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":71.4,"speedDeltaPct":-83.8,"contextDeltaTokens":268928,"parityGaps":["No extended-thinking mode","Loses the free tier"],"parityWins":["Context grows from 131,072 to 400,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":0.6666666666666666,"closeness":0.7153219909635264,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-6-flash","name":"Gemini 3.6 Flash","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":1185.7,"speedDeltaPct":-85.4,"contextDeltaTokens":868928,"parityGaps":[],"parityWins":["Context grows from 131,072 to 1,000,000 tokens","Max output grows from 32,768 to 64,000 tokens"],"parityScore":1,"closeness":0.7114618337884244,"scoredOn":["parity","effort","price","speed"]}]},{"slug":"deepseek-v4-flash","name":"DeepSeek V4 Flash","provider":"DeepSeek","blendedPerM":0.175,"stayPut":false,"picks":{"closest":"deepseek-v4-pro","cheapest":"gpt-oss-20b","fastest":"cerebras-gpt-oss-120b","sameProvider":"deepseek-v4-pro"},"alternatives":[{"slug":"deepseek-v4-pro","name":"DeepSeek V4 Pro","provider":"DeepSeek","effort":"drop-in","effortReason":"Same provider — change the model string, nothing else.","priceDeltaPct":210.7,"speedDeltaPct":-48.5,"contextDeltaTokens":0,"parityGaps":[],"parityWins":["Max output grows from 32,000 to 64,000 tokens"],"parityScore":1,"closeness":0.8950683594058091,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-luna","name":"GPT-5.6 Luna","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":157.1,"speedDeltaPct":-4.5,"contextDeltaTokens":272000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 400,000 tokens","Max output grows from 32,000 to 64,000 tokens"],"parityScore":1,"closeness":0.848655417172438,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-gpt-oss-120b","name":"GPT-OSS 120B (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":157.1,"speedDeltaPct":1756.1,"contextDeltaTokens":3072,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 131,072 tokens","Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.6666666666666666,"closeness":0.8117933167158171,"scoredOn":["parity","effort","price","speed"]},{"slug":"gpt-5-6-terra","name":"GPT-5.6 Terra","provider":"OpenAI","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":2471.4,"speedDeltaPct":-40.9,"contextDeltaTokens":272000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 400,000 tokens","Max output grows from 32,000 to 128,000 tokens"],"parityScore":1,"closeness":0.7958801951780667,"scoredOn":["parity","effort","price","speed"]},{"slug":"cerebras-glm-4-7","name":"GLM 4.7 (Cerebras)","provider":"Cerebras","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":1257.1,"speedDeltaPct":1400,"contextDeltaTokens":72000,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 200,000 tokens","Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.6666666666666666,"closeness":0.768144630695499,"scoredOn":["parity","effort","price","speed"]},{"slug":"glm-5-2","name":"GLM-5.2","provider":"Z.ai","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":1128.6,"speedDeltaPct":null,"contextDeltaTokens":872000,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 1,000,000 tokens","Max output grows from 32,000 to 64,000 tokens"],"parityScore":0.6666666666666666,"closeness":0.7671961387029494,"scoredOn":["parity","effort","price"]},{"slug":"gpt-oss-20b","name":"GPT-OSS 20B","provider":"Groq","effort":"config","effortReason":"Keep the `openai` SDK; change `baseURL` and the API key.","priceDeltaPct":-25,"speedDeltaPct":748.5,"contextDeltaTokens":3072,"parityGaps":["Loses prompt caching"],"parityWins":["Context grows from 128,000 to 131,072 tokens","Max output grows from 32,000 to 32,768 tokens"],"parityScore":0.6666666666666666,"closeness":0.760578022044259,"scoredOn":["parity","effort","price","speed"]},{"slug":"gemini-3-5-flash-lite","name":"Gemini 3.5 Flash Lite","provider":"Google","effort":"code-change","effortReason":"contents/parts structure instead of messages; an OpenAI-compatible endpoint exists but only covers a subset of parameters.","priceDeltaPct":221.4,"speedDeltaPct":22.7,"contextDeltaTokens":872000,"parityGaps":[],"parityWins":["Context grows from 128,000 to 1,000,000 tokens"],"parityScore":1,"closeness":0.7487349483391782,"scoredOn":["parity","effort","price","speed"]}]}]}