Compare Start here

OpenAILanguage model

GPT-5.6 Terra

OpenAI language model.

Compare head-to-head Race it in 3D

Overall Intelligence

#32

of 61
39.5 pts behind #1

0

races led

0

weeks at #1

51.6

score

Released: Jul 9, 2026; 3 mo ago
Context: 1.05M; 128K max output
Price: $2 / $12; per 1M in/out · $4.50 blended · OpenRouter
Speed: 159 tok/s; 0.81s to first token (AA)
Modalities: image · text; outputs text
Weights: Closed; API and apps

The 30-second version

OpenAI language model.

Key takeaways

1OpenAI language model.

Where it ranks

Latest real results · click any race to watch it

5.6 TerraBest on each axis

Overall

32Overall+39.5 pts

Coding

34Coding+58.6 pts 41Frontend & UI+286 Elo 32Coding Q&A+62 Elo 40React Web Apps+342 Elo 29Agentic Coding+20.1 pts 15Terminal & DevOps+28.2 pts

Hard Reasoning

27Hard Reasoning+44.0 pts 29Frontier Questions+18.5 pts 13ARC-AGI-2+11.1 pts 27Common-Sense Reasoning+33.0 pts

Math

15Math+24.0 pts 11Competition Math+0.3 pts 10Research Math+7.7 pts 36Math Chat+72 Elo

Science

36Science+51.6 pts 14Graduate Science QA+3.1 pts 52Medicine & Health+124 Elo 27Scientific Coding+11.9 pts 41Science Q&A+94 Elo

Writing & Language

39Writing & Language+54.3 pts 43Creative Writing+122 Elo 41Professional Writing+103 Elo 33Multilingual+82 Elo 34Instruction Following+80 Elo 39Conversation+104 Elo

Multimodal

22Multimodal+45.9 pts 31Image Understanding+54 Elo 10Documents & Charts+45 Elo

Tool Use & Agents

14Tool Use+11.1 pts

Long Context

14Long Context+19.6 pts 14Long-Document Reasoning+5.7 pts

Speed & Cost

10Speed & Cost+33.2 pts 6Output Speed+231 tok/s 11Time to First Token+0.32s 36Intelligence per Dollar+608.7 pts

Trust & Factuality

38Trust & Factuality+72.3 pts 35Factuality+32.8 pts

Overall Intelligence over time

#1 frontier5.6 TerraGrok 4.5V4 Pro

Update history

The distilled version first; open any release for the deeper breakdown.

60-second briefing

5.6 Terra, explained

Animated summary · captions from the auto-written script

From OpenAI's own guide, distilled

Prompting guides for GPT-5 and later models

Reasoning models do their own step-by-step thinking; give goals and constraints rather than scripting the chain of thought.
Use the reasoning-effort and verbosity controls instead of prompt tricks to trade depth for speed.
Contradictory instructions waste reasoning tokens; audit system prompts for conflicts.
For agents, define stop conditions and how eager the model should be to keep going without asking.
Structured outputs (JSON schema) beat asking politely for a format.

All official guides, distilled →

About OpenAI

Maker of ChatGPT and the GPT series, pursuing general-purpose AI at consumer scale.

Optimising for

Frontier reasoning and math, broad consumer usefulness inside ChatGPT, and tiered models (Astra, Sol, Luna) that trade depth for price and speed.

San Francisco, US · founded 2015

Closest rivals

5.6 Terra is 32nd in overall intelligence; leader is Claude Opus 5.5.