GPT-6 Astra
Second on WebDev and Agent boards, but GPT-6.1 Sol is close for a fifth of the price; middling at creative writing.
Should you use it?
Pick it if
Teams on the hardest coding and agent jobs where a small quality lead is worth paying for.
Best for: The hardest coding and agent jobs, where a small lead is worth paying for (Chat models Top 20)
Skip it if
Cost matters at all; GPT-6.1 Sol or Claude Opus 5.5 do the same jobs for less.
You give up: Five times GPT-6.1 Sol's token price; middling at creative writing
The short version
- What this is
- OpenAI's most capable model, launched 3 Sep 2026, in ChatGPT Plus and Pro, Codex and the API.
- Why it matters
- Knowing that GPT-6.1 Sol gets close on the same boards for much less can save heavy API users a lot.
- Second on arena.ai WebDev and Agent; only mid-table on creative writing s1 s3 s7.
- API costs $10 / $50 per 1M tokens, five times GPT-6.1 Sol s11.
- In ChatGPT from Plus at $20/month s9 s10.
Good at
Our take
Astra, launched on 3 Sep 2026 s17, is among the strongest models for code and agent work: second on both arena.ai WebDev and the Agent board on 11 Oct 2026 s1 s3. OpenAI calls it "the world's most intelligent and aligned model" s16; on the public boards we use, Claude Opus 5.5 ranks above it on both.
The question is price. GPT-6.1 Sol sits two to three places lower on the same boards and scores 52 against Astra's 53 on the Artificial Analysis index, at about a fifth of the cost per task s5. Unless you are already hitting Sol's limits, start there. Astra's writing and Expert ranks are mid-table s6 s7, so it scores 80 for writing and 89 for research, against 90 for coding.
Scores by job
A score from 0 to 100 per job, against everything available today. Overall 84 across 3 jobs, the average of the job scores.
New to this? What the scores and labels mean
- 0 to 100
- Our score for one job, like writing or coding, compared with everything else out today. 100 is the best available today. For AI models we build it from public data, such as Arena's blind votes and Epoch AI's independent tests; apps and tools are scored on our own checklist. The same tool can score differently for different jobs, and where two likely ranges overlap, we call it a tie.
- Our verdict
- Use it: a top choice for most people. Worth trying: good, and worth a go. Depends: good for some people or one kind of job. Wait: too early to tell, or not out for everyone yet. Skip: something else does the job better.
- First look
- Just released. We say what it is and whether to bother, and give it a score once public results exist.
- No score yet
- None of the public sources we use covers it yet, or the only evidence is from a source whose numbers we can't use, so we describe it in words instead.
- Best buy
- Nothing cheaper scores as well for that job.
- Best free
- The best score for that job that costs nothing, through a free plan or by running on your own computer.
- Pay as you go
- No monthly plan: you pay for each use, usually through a developer account.
- Checked
- The day we last confirmed the score and the price against the sources. Prices change often, so check the date.
- AI model
- The engine underneath, like GPT or Claude. You use it through an app or, if you code, directly.
- Open source
- Free to download and run on your own computer, so nothing leaves your machine.
#5 of the 15 we have scored for this job.
From blind votes (Arena) and independent tests (Epoch AI) · Arena data, CC BY 4.0, Epoch AI data, CC BY 4.0
2 sourcesData read 20 Oct
#7 of the 12 we have scored for this job.
From blind votes (Arena) · Arena data, CC BY 4.0
1 source so farData read 20 Oct
#9 of the 21 we have scored for this job.
From blind votes (Arena) and independent tests (Epoch AI) · Arena data, CC BY 4.0, Epoch AI data, CC BY 4.0
1 source so farData read 20 Oct
#2 in Arena's blind votes for agents. Those votes give a rank but no rating, so there is no score yet.
#2 in Arena's blind votes for agents (rank only, no score yet) · Arena data, CC BY 4.0
2 sourcesData read 20 Oct
Alternatives
Better or cheaper for agents, from our scores, then the ones we'd also look at. The same kind of product comes first.
- Claude Opus 5.5BetterText model#1$20 a monthRanked #1 on Arena's Agent board in agents, against #2 here; $20 a month. First on WebDev and Agent at $4 / $20 per 1M tokens, and stronger at writing. Side by side
- Gemini 4 ArgonCheaperText model#6$4.99 a monthRanked #6 on Arena's Agent board in agents for $4.99 a month, against $20 a month here. Side by side
- MiMo-V2.6-ProFreeText model#13FreeRanked #13 on Arena's Agent board in agents, with a free plan. Side by side
- GPT-6.1 SolText model#4$20 a monthTwo places lower on WebDev and three on Agent, at a fifth of the token price. Side by side
Value for money
Where it sits in agents: higher is a better score, further left is cheaper. Green dots are the Best buys.
Evidence
Everything behind the call and the scores, folded so the page stays short. Spotted a wrong price, score or link? Report a mistake.
How we scored this
The research, writing and coding scores are our own, built from public data: blind votes on Arena and, where they cover the job, Epoch AI's independent tests. How we combine them is our method and we don't publish it. For agents, Arena's blind votes give a rank but no rating, so we show the rank and no score yet.
Public leaderboardIndependent test
Reviews are drafted with help from AI models, including Anthropic's Claude. Model scores come only from outside leaderboards and independent evaluations, by a written rule.
No one paid for this review or gave us free access, and we earn nothing if you sign up. How we score
Sources 17
- s1arena.ai Code Arena: WebDev leaderboard
- s2arena.ai Text: Coding leaderboard
- s3arena.ai Agent leaderboard
- s4arena.ai Agent Pareto and cost per task
- s5Artificial Analysis LLM leaderboard
- s6arena.ai Text: Expert leaderboard
- s7arena.ai Text: Creative Writing leaderboard
- s8arena.ai Text: Instruction Following leaderboard
- s9OpenAI Help: What is ChatGPT Plus?
- s10ChatGPT pricing and plan comparison
- s11OpenAI API pricing
- s12OpenAI Help: How your data is used to improve model performance
- s13OpenAI API docs: Your data
- s14arena.ai Agent Alignment Index
- s15OpenAI API docs: Models
- s16OpenAI: Introducing GPT-6 Astra
- s17Techmeme: OpenAI launches GPT-6 Astra
Something here out of date or wrong? Report a mistake and we will check it against the source.
History 4 changes
Every change to a rating, the price or our call, with the reason. Scores are against what exists on the day, so they drift down as better things ship. Older entries come from before we moved to scores.
- OpenAI's ChatGPT release notes (16 Oct 2026): for Pro users, rolling out, GPT-6 Pro in chat responds faster, uses tools better and can run several agents in parallel on parts of a task. No rating change: ratings use the best model each plan lets you choose, which is unchanged.
- New: Coding ratedNew: Agents ratedNew: Research ratedNew: Writing ratedReviewed: coding, agents, research, writing.
- Audit: arena.ai Agent median cost per task corrected from $3.43 to $3.42 to match the main table; $3.43 is the Pareto view's 14-day median. Ratings re-derived, unchanged.
- Audit: applied our tie rule to the Artificial Analysis rank, now 5th instead of 4th. Still the same, so agents stays the same and research stays the same.
Use it well 4 guides on this site
Pages on this site that help with the jobs above.