MDL
GPT-5 mini
OpenAI
Verified against Artificial Analysis · Jul 13, 2026
Capability
82
percentile index
Price /M
$0.25 / $2
in / out
Context
400K
tokens
Avg score
61.0
10 benchmarks
AA Intelligence
30.9
index
Speed
107 tok/s
TTFT 15s
Released
Aug 7, 2025
About
GPT-5 mini is a proprietary model in the GPT family from OpenAI. Benchmarked on 10 evals, averaging 61.0.
Benchmark scores · 10
avg 61.0 — higher is betterThe research behind this model
all papers mentioning it →FineVerify: Scaling Test-Time Compute with Fine-Grained Self-Verification for Agentic Search▲ 8 on HF · May 30, 2026τ-Rec: A Verifiable Benchmark for Agentic Recommender Systems▲ 1 on HF · Jun 8, 2026A Dataset of Robot-Patient and Doctor-Patient Medical Dialogues for Spoken Language Processing TasksMay 26, 2026LogDx-CI: Benchmarking Log Reduction Tools for LLM Root-Cause DiagnosisMay 26, 2026Memory Architectures for Multi-Turn Text-to-SQL: A Benchmark and Empirical StudyMay 25, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed107 tok/s
Latency (TTFT)15s
Price in / out$0.25 / $2 per 1M
Blended price$0.688 / 1M
Math index85.0
Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →
History
full ledger →Capability over time
AA Long-Context Reasoning68 → 66
13d ago
Terminal-Bench Hard33.3 → 28.8
13d ago
τ²-bench68.4 → 71.1
13d ago
IFBench75.4 → 71.2
13d ago
AIME90.7 → 85
13d ago
SciCode39.2 → 41
13d ago
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$2/M