GPT-6.1 Sol vs GPT-6 Astra
Wins 2 of 8 areas
Math · Long documents
Wins 4 of 8 areas
Coding · Reasoning · Facts · Following instructions
GPT-6 Astra wins more areas, narrowly.GPT-6.1 Sol is cheaper and better at math.
Both rank among the ten best models we track in reasoning, facts and math.
Scores updated · 44 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.
- CodingWriting and fixing software03GPT-6 Astra3 of 4 tests · 1 tie
- ReasoningHard problems that need careful thinking02GPT-6 Astra2 of 6 tests · 4 ties
- FactsGetting facts right instead of making them up02GPT-6 Astra2 of 3 tests · 1 tie
- Following instructionsDoing exactly what it is asked01GPT-6 Astra1 of 1 test
- MathCompetition and research-level math10GPT-6.1 Sol1 of 3 tests · 2 ties
- Long documentsFinding answers in very long texts10GPT-6.1 Sol1 of 1 test
- AgentsCarrying out multi-step tasks on its own11Even1 each
- Images and chartsUnderstanding pictures, charts and video00Even0 each · 2 ties
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.
GPT-6.1 Sol 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
- Solves new puzzle games it has never seenARC-AGI-3+47.2points ahead
- Complex command-line tasks across many fieldsTerminal-Bench 4.0+3points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+3points ahead
Where GPT-6.1 Sol pulls ahead
- Extremely hard research-level math problemsFrontierMath Tier 4+2.4points ahead
- Reasons across sets of long documentsAA-LCR+2.3points ahead
- Real work tasks from 44 professionsGDPVal+1.1points ahead
Every test, side by side
All 44 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-6 Astra
- LiveBench · Agentic CodingGPT-6 Astra by 2.754.557.3+2.7
- SciCodeGPT-6 Astra by 2.354.256.5+2.3
- LMArena · WebDevGPT-6 Astra by 21 rating points17431764+21 rating
- LiveBench · Codingtie80.480.4tie
AgentsEven
- Terminal-Bench 4.0GPT-6 Astra by 356.159.1+3
- GDPValGPT-6.1 Sol by 1.154.753.6+1.1
ReasoningGPT-6 Astra
- ARC-AGI-3GPT-6 Astra by 47.252.799.9+47.2
- Humanity's Last ExamGPT-6 Astra by 1.852.954.7+1.8
- ARC-AGI-2tie94.295tie
- SimpleBenchtie82.983.6tie
- LiveBench · Reasoningtie92.692.7tie
- CritPttie31.731.7tie
FactsGPT-6 Astra
- AA-Omniscience · Non-hallucinationGPT-6 Astra by 345.748.7+3
- SimpleQA VerifiedGPT-6 Astra by 1.773.975.6+1.7
- AA-Omniscience · Accuracytie62.162.6tie
Images and chartsEven
- MMMU-Protie8686.9tie
- LMArena · Visiontie12861279tie
MathGPT-6.1 Sol
- FrontierMath Tier 4GPT-6.1 Sol by 2.410097.6+2.4
- FrontierMath Tiers 1-3 (v2)tie93.793.7tie
- LiveBench · Mathematicstie96.896.8tie
Following instructionsGPT-6 Astra
- LiveBench · Instruction FollowingGPT-6 Astra by 1.474.275.6+1.4
Other results22 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.
- ARC-AGI-3 (Standard harness, max effort)GPT-6 Astra by 1052.762.7+10
- Terminal-Bench-Science 0.1GPT-6 Astra by 5.258.163.3+5.2
- HLE-DiamondGPT-6 Astra by 558.863.8+5
- HiL-Bench (Tools-allowed)GPT-6 Astra by 362.365.3+3
- GDPval-AA v2.1 (max effort)GPT-6.1 Sol by 25 rating points16091584+25 rating
- ARC-AGI-3 (Provider Adapter harness, max effort)GPT-6 Astra by 2.496.298.6+2.4
- AA-OmniscienceGPT-6 Astra by 1.941.543.4+1.9
- OSWorld 2.0GPT-6 Astra by 1.271.472.6+1.2
- DeepSWE 1.1GPT-6.1 Sol by 1.175.274.1+1.1
- ARC-AGI-1GPT-6 Astra by 196.597.5+1
- CritPt v2 HardGPT-6 Astra by 169.570.5+1
- AA Intelligencetie51.852.7tie
- HealthBench Professionaltie65.166tie
- livebench_languagetie90.189.4tie
- HealthBench Consensus length-adjustedtie9695.5tie
- HealthBench Professional length-adjustedtie64.264.7tie
- Terminal-Bench 4.0tie58.257.7tie
- HealthBench Hard length-adjustedtie36.236.6tie
- HealthBenchtie58.558.2tie
- livebench_data_analysistie82.783tie
- HealthBench length-adjustedtie58.558.3tie
- GDP (Surge AI)tie3131tie
Questions people ask
Which is better, GPT-6.1 Sol or GPT-6 Astra?
GPT-6 Astra wins four of the eight areas we test: coding, reasoning, facts and following instructions. GPT-6.1 Sol wins math and long documents, and costs 80% less. They are level on agents and images and charts.
Which is better for coding?
GPT-6 Astra. It wins 3 of the 4 coding tests both models report; GPT-6.1 Sol wins none, and 1 is a tie.
Which is cheaper?
GPT-6.1 Sol 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 GPT-6.1 Sol about 80% cheaper for the same work.
How do you compare the two?
We use the 44 benchmark tests both models have published scores on. The verdict counts the 22 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 22 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.