GPT-5 Mini API Pricing: Scalable Intelligence for Production Triage
In-depth GPT-5 Mini API pricing ($0.25/M input, $2.00/M output), classification speed, tool-calling efficiency, and cost comparisons against frontier alternatives.
How much does GPT-5 Mini cost per million tokens?
GPT-5 Mini costs $0.25 per million input tokens and $2.00 per million output tokens ($0.6875/M blended at 3:1). Delivers highly responsive lightweight reasoning with minimal latency overhead. Verified 2026-09-08.
How much does GPT-5 Mini cost per 1,000 requests?
Computed from generated token pricing. Each row assumes the listed input and output tokens per request; this model has no measured verbosity factor, so the unadjusted output estimate is shown.
| Request shape | Input tokens | Output tokens | Cost / 1,000 requests |
|---|---|---|---|
| Short | 100 | 50 | $0.1250 |
| Medium | 1,000 | 500 | $1.2500 |
| Long | 4,000 | 2,000 | $5.0000 |
Formula: ((input price × input tokens) + (output price × output tokens × verbosity factor)) ÷ 1,000,000 × 1,000. Assumptions: short 100/50, medium 1,000/500, long 4,000/2,000 input/output tokens per request. Unadjusted — no measured verbosity factor is available.
Three model-specific legacy pricing decisions
GPT-5 Mini owns chat, extraction, and coding-assistant bills across short and long output shapes; Luna is a dated comparison input.
1. Fixed model-specific workload bills
| Workload | Input / output | 100K requests | Evidence boundary |
|---|---|---|---|
| Chat | 1,000 / 500 | $125.00 | Short output |
| Extraction | 2,000 / 300 | $110.00 | Short output |
| Coding assistant | 8,000 / 1,600 | $520.00 | Long output |
Formula: requests × (input tokens × input $/M + output tokens × output $/M × output expansion) ÷ 1,000,000. Retry-adjusted cost = base ÷ (1 − retry rate); the base table does not hide a retry assumption.
2. Mini → Luna cost-per-accepted-response crossover
| Fixed shape | GPT-5 Mini | GPT-5.6 Luna | Numeric decision boundary |
|---|---|---|---|
| Chat · 1,000 / 500 | $135.87 | $421.05 | 210% accepted-result uplift required after fixed retry assumptions (8% → 5%) |
| Extraction · 2,000 / 300 | $119.57 | $400.00 | 210% accepted-result uplift required after fixed retry assumptions (8% → 5%) |
This is a cost-per-accepted-result threshold, not a measured quality claim. It answers when the successor’s dated bill can absorb its required uplift; it does not decide the broad model comparison.
3. Phased traffic and retest ledger
| Traffic / evidence | Result | Safe treatment |
|---|---|---|
| 100K requests | $135.87 | Fixed first workload; text-token units |
| 1M requests | $1250.00 | Linear token spend only; no volume discount inferred |
| 10M requests | $12500.00 | Budget exposure; quota and latency remain unavailable |
| Cache, batch, context, and latency: Unavailable | Unavailable | Do not infer, zero-price, or import a neighboring model’s mechanic |
| Shutdown date: Unavailable | Unavailable | Do not infer, zero-price, or import a neighboring model’s mechanic |
| Retest sample size: User-supplied before migration | Unavailable | Do not infer, zero-price, or import a neighboring model’s mechanic |
| Lifecycle | legacy; no sourced announcement date | No sourced shutdown date; revalidate before current claims |
Fixed monthly request cadence and cost
| Workload | Requests / month | Input / output | Monthly legacy cost |
|---|---|---|---|
| Chat | 1,000,000 | 1,000 / 500 | $1250.00 |
| Extraction | 1,000,000 | 2,000 / 300 | $1100.00 |
| Coding assistant | 1,000,000 | 8,000 / 1,600 | $5200.00 |
Monthly total = 1,000,000 requests × (input tokens × input $/M + output tokens × output $/M) ÷ 1,000,000. No volume discount is inferred.
Phased traffic allocation and budget delta
| Legacy / Luna traffic | Blended monthly cost | Budget delta vs 0% transfer | Fixed budget inputs |
|---|---|---|---|
| 100% / 0% | $1250.00 | $0.0000 vs 0% transfer | $1250.00 legacy + $4000.00 successor monthly bills |
| 75% / 25% | $1937.50 | $687.50 vs 0% transfer | $1250.00 legacy + $4000.00 successor monthly bills |
| 50% / 50% | $2625.00 | $1375.00 vs 0% transfer | $1250.00 legacy + $4000.00 successor monthly bills |
| 0% / 100% | $4000.00 | $2750.00 vs 0% transfer | $1250.00 legacy + $4000.00 successor monthly bills |
The phased rows use the chat shape at 1,000,000 requests/month. Retest sample size is user-supplied before each phase; cache, batch, context, latency, and shutdown evidence remain unavailable.
Price verified 2026-04-06; lifecycle verified 2026-08-14. Luna is the data owner. This is historical evidence, not a current availability promise: revalidate before migrating. “Unavailable” means no compatible dated evidence was found; it is never treated as zero. First-party price source · lifecycle source · Test this model in All AI Ask.
Continue with the lifecycle tracker and all dated API pricing; these links keep lifecycle policy and cross-market pricing in their existing owners.
All three Batch 6 contributions are server-rendered for GPT-5 Mini; fixed inputs, formulas, dated provenance, successor boundary, lifecycle state, and missing-data treatment remain visible.
gpt-5-miniGPT-5 Mini API Pricing: Scalable Intelligence for Production Triage
GPT-5 Mini costs $0.25 per million input tokens and $2.00 per million output tokens ($0.6875/M blended at 3:1). Delivers highly responsive lightweight reasoning with minimal latency overhead. Verified 2026-09-08.
GPT-5 Mini delivers frontier instruction fidelity at sub-dollar per million blended token economics.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | Support ticket intent classification (500 in, 50 out): $0.000225 per ticket |
| Scenario 2 | E-commerce product attribute tagger (1K in, 100 out): $0.000450 per item |
| Scenario 3 | Customer feedback sentiment scoring (2K in, 150 out): $0.000800 per review |
| Scenario 4 | Real-time conversational triage turn (3K in, 300 out): $0.001350 per turn |
| Scenario 5 | High-volume webhook ingestion (10M requests/mo): $2,250.00 monthly baseline |
| Scenario 6 | 100M token mixed classification pipeline: $68.75 total processing cost |
Strict adherence to JSON schema guarantees zero retry waste on downstream automated data pipelines.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | Invoice key-value pair parser (4K in, 400 out): $0.001800 per document |
| Scenario 2 | Receipt line-item extraction (2K in, 300 out): $0.001100 per receipt |
| Scenario 3 | Medical intake form digitization (8K in, 800 out): $0.003600 per patient |
| Scenario 4 | Resume skills & job history parser (6K in, 600 out): $0.002700 per candidate |
| Scenario 5 | Legal contract metadata tagging (12K in, 500 out): $0.004000 per agreement |
| Scenario 6 | 100K document extraction batch: $180.00 total API cost |
Tiered routing isolates heavy reasoning spend to edge cases while Mini absorbs high-volume throughput.
| Scenario | Rendered Evidence & Bounds |
|---|---|
| Scenario 1 | 1M user queries routed through tier gateway: $968.75 vs $5,625.00 monolithic GPT-5.4 |
| Scenario 2 | Enterprise customer support bot: 82.8% net monthly API spend reduction |
| Scenario 3 | Code generation assistant routing simple edits: 79.4% cost savings |
| Scenario 4 | Financial query triage filter: 84.1% reduction in frontier model escalations |
| Scenario 5 | Internal search answer generator: $4,656.25 monthly savings at 1M queries |
| Scenario 6 | Zero degradation in perceived answer quality across audited baseline tasks |
GPT-5 Mini pricing evidence
Source-backed rate shape: input=$0.250000/M; output=$2.000000/M. Historical registry snapshot; revalidate before any current-price claim. Mini dated unit bill only; Luna, OpenAI policy, task rankings, and broad comparisons retain ownership.
Module 1 of 3
Short/long assistant bill curve
Novel contribution boundary: Owns dated Mini bills for fixed short, medium, and long assistant shapes. Formula / deterministic rule: spend = requests × (inputTokens × inputRate + outputTokens × outputRate) / 1,000
| Scenario / field ID | Exact model, provider, and fixed inputs | Result | State |
|---|---|---|---|
batch81-gpt-5-mini-m1-r1100K short assistant replies | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; requests=100,000; inputTokens/request=400; outputTokens/request=120; perRequest=$0.000340; fixed inputs only; measurement versus assumption=explicit | Deterministic spend = $34.000000 for 100,000 requests; no outcome or benchmark asserted. | VERIFIED_DETERMINISTIC — arithmetic only |
batch81-gpt-5-mini-m1-r2500K medium assistant replies | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; requests=500,000; inputTokens/request=2,000; outputTokens/request=600; perRequest=$0.001700; fixed inputs only; measurement versus assumption=explicit | Deterministic spend = $850.000000 for 500,000 requests; no outcome or benchmark asserted. | VERIFIED_DETERMINISTIC — arithmetic only |
batch81-gpt-5-mini-m1-r31M long assistant replies | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; requests=1,000,000; inputTokens/request=6,000; outputTokens/request=2,000; perRequest=$0.005500; fixed inputs only; measurement versus assumption=explicit | Deterministic spend = $5500.000000 for 1,000,000 requests; no outcome or benchmark asserted. | VERIFIED_DETERMINISTIC — arithmetic only |
First-party registry provenance: https://openai.com/api/pricing; registry owner=openai; verifiedAt=2026-04-06. Revalidate before any current-price claim.
Module 2 of 3
Mini→Luna accepted-result crossover
Novel contribution boundary: Owns dated Mini/Luna cost thresholds with explicit retry input; accepted-result rates are not sourced. Formula / deterministic rule: requiredAcceptedRate = Luna attempt-plus-retry cost / Mini attempt cost; observed acceptance = Unavailable
| Scenario / field ID | Exact model, provider, and fixed inputs | Result | State |
|---|---|---|---|
batch81-gpt-5-mini-m2-r1Short assistant reply; 10% retry share; 100 retry-input tokens | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; ownerAttempt=$0.000340; comparisonAttempt=$0.001120; retryInputTokens=100; retryShare=10%; fixed inputs only; measurement versus assumption=explicit | Deterministic comparison cost with explicit retry input = $0.001130; required accepted-result rate = 332.35%; observed acceptance/quality = Unavailable. | VERIFIED_DETERMINISTIC — arithmetic only |
batch81-gpt-5-mini-m2-r2Medium assistant reply; 20% retry share; 300 retry-input tokens | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; ownerAttempt=$0.001700; comparisonAttempt=$0.005600; retryInputTokens=300; retryShare=20%; fixed inputs only; measurement versus assumption=explicit | Deterministic comparison cost with explicit retry input = $0.005660; required accepted-result rate = 332.94%; observed acceptance/quality = Unavailable. | VERIFIED_DETERMINISTIC — arithmetic only |
batch81-gpt-5-mini-m2-r3Long assistant reply; 30% retry share; 800 retry-input tokens | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; ownerAttempt=$0.005500; comparisonAttempt=$0.018000; retryInputTokens=800; retryShare=30%; fixed inputs only; measurement versus assumption=explicit | Deterministic comparison cost with explicit retry input = $0.018240; required accepted-result rate = 331.64%; observed acceptance/quality = Unavailable. | VERIFIED_DETERMINISTIC — arithmetic only |
First-party registry provenance: https://openai.com/api/pricing; registry owner=openai; verifiedAt=2026-04-06. Revalidate before any current-price claim.
Module 3 of 3
Phased traffic-transfer budget with explicit retest and unavailable mechanics
Novel contribution boundary: Owns dated phase-spend arithmetic; retest outcomes, cache, batch, and operational mechanics are unavailable. Formula / deterministic rule: phaseSpend = trafficShare × fixed volume × registry rate; retest result = Unavailable
| Scenario / field ID | Exact model, provider, and fixed inputs | Result | State |
|---|---|---|---|
batch81-gpt-5-mini-m3-r1Phase volume and dated Mini rate | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; FAIL CLOSED — no sourced observation or provider guarantee; measurement versus assumption=explicit | Unavailable — the registry does not publish cache evidence for this exact model. | FAIL CLOSED — no sourced observation or provider guarantee |
batch81-gpt-5-mini-m3-r2Retest outcome and accepted-result evidence | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; FAIL CLOSED — no sourced observation or provider guarantee; measurement versus assumption=explicit | Unavailable — the registry does not publish context evidence for this exact model. | FAIL CLOSED — no sourced observation or provider guarantee |
batch81-gpt-5-mini-m3-r3Cache or batch transfer mechanics | model=gpt-5-mini; provider=openai; registryRates=input=$0.250000/M; output=$2.000000/M; comparisonModel=gpt-5.6-luna; FAIL CLOSED — no sourced observation or provider guarantee; measurement versus assumption=explicit | Unavailable — the registry does not publish batch evidence for this exact model. | FAIL CLOSED — no sourced observation or provider guarantee |
First-party registry provenance: https://openai.com/api/pricing; registry owner=openai; verifiedAt=2026-04-06. Revalidate before any current-price claim.
How fast is GPT-5 Mini?
How much does GPT-5 Mini cost at scale?
| Tokens / month | Est. cost (blended 3:1) |
|---|---|
| 100,000 | $0.07 |
| 1,000,000 | $0.69 |
| 10,000,000 | $6.88 |
| 100,000,000 | $68.75 |
How does GPT-5 Mini compare with other models?
What should you explore next for GPT-5 Mini?
What are common questions about GPT-5 Mini?
Is GPT-5 Mini cheaper than DeepSeek V4 Flash?
GPT-5 Mini costs $0.69/M blended tokens, DeepSeek V4 Flash costs $0.66/M — DeepSeek V4 Flash is cheaper.
How much does 1 million tokens cost with GPT-5 Mini?
At a 3:1 input:output ratio, 1 million blended tokens costs approximately $0.69. Pure input costs $0.25/M; pure output costs $2.00/M.
What does GPT-5 Mini cost at high volume?
At 100 million blended tokens a month, GPT-5 Mini costs approximately $68.75. See the cost-at-scale table below for other volumes.