MDL
GPT-4.1
OpenAI
Verified against Artificial Analysis · Jul 13, 2026
Capability
63
percentile index
Price /M
$2 / $8
in / out
Context
1M
tokens
Avg score
50.4
13 benchmarks
AA Intelligence
19.4
index
Speed
128 tok/s
TTFT 0.7s
Released
Apr 14, 2025
About
GPT-4.1 is a proprietary model in the GPT-4.1 family from OpenAI. Benchmarked on 13 evals, averaging 50.4.
proprietarytextvision
Benchmark scores · 13
avg 50.4 — higher is betterThe research behind this model
all papers mentioning it →Online Skill Learning for Web Agents via State-Grounded Dynamic Retrieval▲ 10 on HF · Jun 3, 2026GRASP: Gated Regression-Aware Skill Proposer for Self-Improving LLM AgentsMay 28, 2026Casual as an Anchor: Resolving Supervision Misalignment in Formality Transfer DatasetMay 28, 2026BenchTrace: A Benchmark for Testing Reflection Ability and Controlled Evolution in LLM AgentsMay 28, 2026Let the Results Speak: A Replication-First Paradigm for LLM Behavioral BenchmarkingMay 27, 2026
Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.
Performance & price
Output speed128 tok/s
Latency (TTFT)0.7s
Price in / out$2 / $8 per 1M
Blended price$3.5 / 1M
Math index34.7
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
Append-only ledger — every observed change to this model's numbers.
API providers
Artificial Analysis$8/M
OpenAI1M ctx · $8/M