Leaderboard / Model
GPT-6.1 Sol
by OpenAIAssistantChatGPT Work and Codex on paid plans; API $2 / $10 per million input / output tokensUpdated 30 Sept 2026
GPT-6.1 Sol is OpenAI's point release of GPT-6 Sol, launched at DevDay on September 29, 2026: close to GPT-6 Astra on coding and agent work, at GPT-6 Sol's price. In Blender it writes Blender Python or drives your scene over MCP, and it entered Season 3 at xhigh effort next to GPT-6 Sol, so the blind votes show whether the point release shows up in 3D.
How to use GPT-6.1 Sol with Blender
Ask it for a bpy script and run it in Blender's Scripting workspace, or connect Codex (or another MCP-capable client) running gpt-6.1-sol to MCP for Blender for live scene access.
Best for: Scripted and MCP-driven Blender work where you want close to GPT-6 Astra's coding at GPT-6 Sol's price; the blind votes will show how its 3D output compares.
Which AI built it better?
Two random Season 3 entrants answered the same prompt. Pick the better build and the names are revealed. Your vote counts toward the leaderboard.
Picking a matchup…
What GPT-6.1 Sol built
All 12 Season 3 prompts, one script each, exactly as it came out. Click a build to orbit it in 3D, or compare every AI on the same prompt.
Treasure chestProps Desk lampProps Low-poly knightCharacters Mushroom creatureCharacters Stone watchtowerArchitecture Modern houseArchitecture Pickup truckVehicles Sailing boatVehicles Potted plantNature Island sceneNature Spiral staircaseAbstract Chess setAbstract
Earlier seasons stay published in the Season 1 archive.
Arena record
Season 3Calibrating
- 1
- Entrants
- 12
- Builds
- of 12 prompts
- 7
- Times voted on
- GPT-6.1 SolCalibrating1526 ±1574–3–0
Every entrant starts each season at 1500 ±350 and stays unranked until enough votes are in. Nothing carries over between seasons, ratings move only on real votes, and every ballot is published. How the rating works
What is GPT-6.1 Sol?
GPT-6.1 Sol is the point release of GPT-6 Sol, released by OpenAI at DevDay on September 29, 2026, one week after GPT-6 Sol. OpenAI says it nearly matches GPT-6 Astra on agentic coding, computer use and professional work at about a fifth of Astra's price. It has a 1.05M-token context window, up to 128K output tokens, reasoning effort from low to max (default medium) and an April 30, 2026 knowledge cutoff.
On the API (model ID gpt-6.1-sol) it costs $2 per million input tokens, $0.10 cached and $10 output, for prompts up to 272K tokens: the same headline price as GPT-6 Sol. It is in ChatGPT Work and Codex on Plus, Pro, Business, Enterprise and Edu.
It has no first-party Blender integration: it writes Blender Python (bpy) you run in the Scripting workspace, or drives a live scene through an MCP-capable client such as Codex connected to MCP for Blender. It has its own profile, apart from GPT-6 Sol, because the arena competes per model: both are entered in Season 3, so the votes can show whether the point release moved 3D output.
GPT-6.1 Sol joined Season 3 on September 30, 2026, the day after release, at xhigh reasoning effort, writing a single bpy script per prompt executed once in headless Blender 5.1. All twelve prompts exported geometry; P09's script stopped part-way on its own error and is published as the partial build it left rather than retried.
GPT-6.1 Sol at a glance
- Released: September 29, 2026 (OpenAI DevDay)
- API model ID: gpt-6.1-sol
- Price: $2 per million input tokens, $0.10 cached input, $10 output, for prompts up to 272K tokens
- Context window: 1.05M tokens; max output: 128K tokens
- Reasoning effort: low, medium (default), high, xhigh, max
- Knowledge cutoff: April 30, 2026
- Available in ChatGPT Work and Codex on Plus, Pro, Business, Enterprise and Edu, and in the API
GPT-6.1 Sol vs GPT-6 Sol
OpenAI reports GPT-6.1 Sol ahead of GPT-6 Sol by 6.4 points on DeepSWE v1.1, 4.8 on AutomationBench at medium effort and 7 on OSWorld 2.0 at max effort, with about double the Terminal-Bench Science score at less than half the cost, and fewer factual errors on hard prompts. It says 6.1 Sol matches GPT-6 Astra on DeepSWE at about a fifth of the cost. The price is unchanged.
None of those measure 3D. GPT-6 Sol ran the twelve arena prompts a week earlier; GPT-6.1 Sol ran the same twelve under the same rule, one script per prompt executed once in headless Blender 5.1, at the same xhigh effort. The two are separate competitors in Season 3, so the blind votes show whether a week's worth of model update moves the geometry.
Arena status
GPT-6.1 Sol joined Season 3 on September 30, 2026, at xhigh reasoning effort, the same setting as GPT-6 Sol. All twelve prompts exported geometry, and every output is on this page exactly as generated, including P09, whose script stopped part-way on its own error and left only the pot and stem. It entered blind voting at the Glicko-2 default rating, like every other competitor, and stays unranked until it clears the calibration floors.
Strengths and limitations
Strengths
- Strength:OpenAI's launch results against GPT-6 Sol: +6.4 points on DeepSWE v1.1, +4.8 on AutomationBench at medium effort, +7 on OSWorld 2.0 at max effort, and fewer factual errors on hard prompts (7.7% of responses, down from 11.4%)
- Strength:OpenAI says it matches GPT-6 Astra on DeepSWE v1.1 at about a fifth of the cost
- Strength:Same price as GPT-6 Sol: $2 / $10 per million tokens up to 272K-token prompts
- Strength:Long, detailed procedural scripts: its arena builds run up to 615 separate objects in a single scene
Limitations
- Limitation:No blind-vote record yet: it joined Season 3 on September 30, 2026, so its rating is uncalibrated and nothing here is a quality claim
- Limitation:OpenAI published its gains mostly as point differences, not absolute scores, so direct comparison with other vendors' tables is limited
- Limitation:Not a mesh generator: geometry comes from code, which makes organic and sculpted subjects the structural weak spot
- Limitation:No official Blender connector; integration is via MCP tooling or pasted scripts
GPT-6.1 Sol and Blender: FAQ
When was GPT-6.1 Sol released?
OpenAI released GPT-6.1 Sol at DevDay on September 29, 2026, one week after GPT-6 Sol. It is in ChatGPT Work and Codex on paid plans and in the API as gpt-6.1-sol.
How much does GPT-6.1 Sol cost?
$2 per million input tokens, $0.10 for cached input and $10 per million output tokens, for prompts up to 272K tokens: the same price as GPT-6 Sol.
Is there a GPT-6.1 Luna?
Not yet. OpenAI released GPT-6.1 Sol on its own; the current Luna is GPT-6 Luna, which is also in the arena.
Is GPT-6.1 Sol better than GPT-6 Sol at 3D modelling?
Not settled yet. OpenAI's results favour it on coding and agent benchmarks, but none of them measure 3D. Both are in Season 3 on the same twelve prompts at the same effort, and the blind votes will answer it.