Cutoff Data
GPT-5 Mini: price, speed and score
Everything we have measured about GPT-5 Mini from OpenAI, taken from real billed API calls rather than a published rate card.
Last reading 9/7/2026, 12:17:06 AM UTC
Input / 1M
—
Output / 1M
—
Context
—
Median response
—
p95 response
—
Uptime, 7 days
0%
Benchmark score
—
Cost per task
—
Checks, 7 days
168
Observed history
- 2026-09-07GPT-5 Mini stopped responding (credit_limit)
- 2026-09-01GPT-5 Mini responded again after a failed check
- 2026-08-13GPT-5 Mini stopped responding (credit_limit)
- 2026-08-11GPT-5 Mini responded again after a failed check
- 2026-08-08GPT-5 Mini stopped responding (credit_limit)
Our coverage of GPT-5 Mini
- 2026-09-06Gemini 3.6 Flash and GPT-5 Mini Top 2026-09-06 Benchmarks at 100%
- 2026-09-05Gemini 3.6 Flash and GPT-5 Mini Hit 100% Benchmark Score on Sept 5, 2026
- 2026-09-04Gemini 3.6 Flash Hits 100% Benchmark Score in Sept 4 Reading
- 2026-09-03Gemini 3.6 Flash Hits Perfect Benchmark Score: Sept 3, 2026 Data
- 2026-09-02Gemini 3.6 Flash and GPT-5 Mini Hit 100% in Sept 2, 2026 Benchmark Tests
- 2026-09-01Gemini 3.6 Flash Hits 100% Benchmark Score on Sept 1, 2026
Compare with
Side-by-side price, speed and score for every model we track sits on the model index, live status board and benchmark table.
How this is measured
Every hour we send one identical prompt to GPT-5 Mini through a single gateway account, record how long the call took and what it was billed, then divide that cost by the tokens consumed to get an observed price per million tokens.
Benchmark figures come from a fixed suite of code-graded tasks re-run against every tracked model. Scores are the share of available points earned; cost is the average spend per task.
These are one buyer's readings on one route. Enterprise pricing, batch discounts and cached-input rates will differ.