Claude Sonnet 5 vs GPT-6 Astra
Wins 1 of 8 areas
Long documents
Wins 7 of 8 areas
Coding · Agents · Reasoning · Facts · Images and charts · Math · Following instructions
GPT-6 Astra is the stronger all-rounder.Claude Sonnet 5 is cheaper and better at long documents.
Both rank among the ten best models we track in agents.
Scores updated · 37 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking05GPT-6 Astra5 of 5 tests
- AgentsCarrying out multi-step tasks on its own04GPT-6 Astra4 of 4 tests
- MathCompetition and research-level math03GPT-6 Astra3 of 3 tests
- CodingWriting and fixing software13GPT-6 Astra3 of 5 tests · 1 tie
- FactsGetting facts right instead of making them up12GPT-6 Astra2 of 3 tests
- Images and chartsUnderstanding pictures, charts and video01GPT-6 Astra1 of 2 tests · 1 tie
- Following instructionsDoing exactly what it is asked01GPT-6 Astra1 of 1 test
- Long documentsFinding answers in very long texts10Claude Sonnet 51 of 1 test
Long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Claude Sonnet 5 costs 80% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-6 Astra pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+68.3points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+45points ahead
- Short factual questions, answered correctlySimpleQA Verified+41.9points ahead
Where Claude Sonnet 5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+11.9points ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+2.1points ahead
- Reasons across sets of long documentsAA-LCR+1.3points ahead
Every test, side by side
All 37 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-6 Astra
- LMArena · WebDevGPT-6 Astra by 233 rating points15411774+233 rating
- Terminal-Bench 2.1GPT-6 Astra by 7.980.588.4+7.9
- SciCodeGPT-6 Astra by 2.254.356.5+2.2
- LiveBench · Agentic CodingClaude Sonnet 5 by 2.159.457.3+2.1
- LiveBench · Codingtie80.780.4tie
AgentsGPT-6 Astra
- Terminal-Bench 4.0GPT-6 Astra by 4514.159.1+45
- BrowseCompGPT-6 Astra by 6.884.791.5+6.8
- GDPValGPT-6 Astra by 5.448.253.6+5.4
- τ-Bench V3 · BankingGPT-6 Astra by 4.137.341.4+4.1
ReasoningGPT-6 Astra
- SimpleBenchGPT-6 Astra by 2360.683.6+23
- CritPtGPT-6 Astra by 14.816.931.7+14.8
- Humanity's Last ExamGPT-6 Astra by 13.441.354.7+13.4
- GPQA DiamondGPT-6 Astra by 591.196.1+5
- LiveBench · ReasoningGPT-6 Astra by 488.792.7+4
FactsGPT-6 Astra
- SimpleQA VerifiedGPT-6 Astra by 41.933.775.6+41.9
- AA-Omniscience · AccuracyGPT-6 Astra by 22.54062.6+22.5
- AA-Omniscience · Non-hallucinationClaude Sonnet 5 by 11.960.648.7+11.9
Images and chartsGPT-6 Astra
- MMMU-ProGPT-6 Astra by 9.677.386.9+9.6
- LMArena · Visiontie12731280tie
MathGPT-6 Astra
- FrontierMath Tier 4GPT-6 Astra by 68.329.397.6+68.3
- FrontierMath Tiers 1-3 (v2)GPT-6 Astra by 28.165.693.7+28.1
- LiveBench · MathematicsGPT-6 Astra by 3.992.996.8+3.9
Following instructionsGPT-6 Astra
- LiveBench · Instruction FollowingGPT-6 Astra by 11.763.975.6+11.7
Other results13 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- Terminal-Bench 4.0GPT-6 Astra by 45.312.457.7+45.3
- AA-OmniscienceGPT-6 Astra by 26.916.443.4+26.9
- DeepSWE 1.1GPT-6 Astra by 20.15474.1+20.1
- LVBenchGPT-6 Astra by 1968.587.5+19
- AA IntelligenceGPT-6 Astra by 14.538.252.7+14.5
- livebench_languageGPT-6 Astra by 14.47589.4+14.4
- livebench_data_analysisGPT-6 Astra by 11.371.783+11.3
- GDPval-AA 2.1GPT-6 Astra by 93 rating points14491542+93 rating
- AA Agentic IndexGPT-6 Astra by 7.244.351.5+7.2
- HealthBench ProfessionalGPT-6 Astra by 5.657.863.4+5.6
- Artificial Analysis Coding IndexGPT-6 Astra by 5.471.576.9+5.4
- Agent's Last Exam (Pass rate)tie33.334.2tie
- HLE (with tools)tie57.457.2tie
Questions people ask
Which is better, Claude Sonnet 5 or GPT-6 Astra?
GPT-6 Astra wins seven of the eight areas we test: coding, agents, reasoning, facts, images and charts, math and following instructions. Claude Sonnet 5 wins long documents, and costs 80% less.
Which is better for coding?
GPT-6 Astra. It wins 3 of the 5 coding tests both models report; Claude Sonnet 5 wins 1, and 1 is a tie.
Which is cheaper?
Claude Sonnet 5 costs $2.00 per million input tokens and $10.00 per million output tokens; GPT-6 Astra costs $10.00 and $50.00. That makes Claude Sonnet 5 about 80% cheaper for the same work.
How do you compare the two?
We use the 37 benchmark tests both models have published scores on. The verdict counts the 24 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 13 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.