LIVE
Introducing Cosmos 3 Edge20/07/26 · Hugging Face|Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling20/07/26 · Anthropic|At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI20/07/26 · NVIDIA|China's open-weights AI strategy is winning20/07/26|Bristol Myers Squibb Building Life Science Industry’s Most Advanced AI Factory on NVIDIA Vera Rubin20/07/26 · NVIDIA|Safety and alignment in an era of long-horizon models20/07/26 · OpenAI|Claude Fable produced a counterexample to the Jacobian Conjecture20/07/26 · Anthropic|Apply for Anthropic’s AI for Science rare disease research grants20/07/26 · Anthropic|AI advice made people less accurate but more confident – sudy19/07/26|Vincentwei1021/video-shotcraft: AI video skill for Claude Code & Codex — cinematic product videos with Remotion: 106 shot recipe cards, 161 motion previews, a production-ready template19/07/26 · Anthropic|Claude Code uses Bun written in Rust now19/07/26 · Anthropic|OpenAI reduces Codex Model Context Size from 372k to 272k19/07/26 · OpenAI|Introducing Cosmos 3 Edge20/07/26 · Hugging Face|Kimi K3, Qwen 3.8, and Anthropic's (Potential) Unravelling20/07/26 · Anthropic|At SIGGRAPH, NVIDIA Advances Graphics and Simulation With Agentic and Physical AI20/07/26 · NVIDIA|China's open-weights AI strategy is winning20/07/26|Bristol Myers Squibb Building Life Science Industry’s Most Advanced AI Factory on NVIDIA Vera Rubin20/07/26 · NVIDIA|Safety and alignment in an era of long-horizon models20/07/26 · OpenAI|Claude Fable produced a counterexample to the Jacobian Conjecture20/07/26 · Anthropic|Apply for Anthropic’s AI for Science rare disease research grants20/07/26 · Anthropic|AI advice made people less accurate but more confident – sudy19/07/26|Vincentwei1021/video-shotcraft: AI video skill for Claude Code & Codex — cinematic product videos with Remotion: 106 shot recipe cards, 161 motion previews, a production-ready template19/07/26 · Anthropic|Claude Code uses Bun written in Rust now19/07/26 · Anthropic|OpenAI reduces Codex Model Context Size from 372k to 272k19/07/26 · OpenAI|
🇺🇸 United States🔬 ReasoningReleased April 2025

o3

OpenAI
RGPD
Verdict

A good performance/compliance balance for a European company.

Overview

o3 uses a chain-of-thought architecture reinforced by RL: it "thinks" for several minutes before answering, making it extraordinarily strong on complex problems. It's the reference model for research, mathematical proofs, and algorithmic competitions. Cost and slowness reserve it for high-end use cases.

Skill profile

FrenchReasoningSpeedCreativitySafety
Arena Elo1410
Verified benchmarks
GPQA87.7%
SWE-Bench71.7%
MATH96.7%
MMLU88.5%
HumanEval96%

What do these scores mean?

GPQA87.7%excellent

PhD-level science questions (physics, chemistry, biology), with no tool access.

SWE-Bench71.7%very good

Resolving real GitHub issues under real conditions (SWE-bench Verified).

MATH96.7%world-class

Competition-level math problems.

MMLU88.5%excellent

General knowledge across dozens of academic subjects.

HumanEval96%world-class

Python code generation from specifications.

Strengths

  • Top-tier reasoning
  • Excellent at code
  • Strong French quality
  • World-class on Arena

Limitations

  • Average speed

Who is it for

A good fit if…
  • you need advanced reasoning or analysis
  • you build with a coding agent
  • you work in French
  • you have GDPR constraints
Skip it if…
  • you need very fast responses

Ideal use cases

  • Advanced math
  • Scientific research
  • Complex algorithms
  • Proofs
  • Competitions

Access & availability

Paid API per tokenConsumer subscription (~200€/mo)

Key specifications

Context
200K
Input price
$2.00 $/M
Output price
$8.00 $/M
Speed
30 tok/s
Price not auditedBenchmarks not audited

Estimate your monthly cost

Per-token API pricing
$117/ month with o3

For the same usage

Indicative estimate based on standard API rates (excluding caching, batch and volume discounts). Always check the official pricing before committing.

Privacy

GDPR-compliantHosting : US/EUAnonymizable data
Advanced data · for experts
Architecture, modalities, detailed cost, full benchmarks
Architecture
Dense + RL
Size
undisclosed
Cutoff date
Nov 2024
Inputs
text, image
Outputs
text, code
License
commercial
Hosting
US/EU
Hallucination score
4/5
Estimated cost (API)
Typical exchange (~3k in / 1k out)
≈ $0.0140
1M in + 1M out
≈ $10.00
All benchmarks
Arena Elo1410
MMLU88.5%
GPQA87.7%
HumanEval96.0%
SWE-Bench71.7%
MATH96.7%

Similar models