GPT-6.2 Sol vs Muse Ember
GPT-6.2 Sol leads Muse Ember on coding, 83.1 to 64.9 on the Vibe Coding Index, and Muse Ember costs 20.0x less per output token ($0.50 vs $10.00 per million). Updated November 22, 2026; 20 benchmarks have a result for both.
GPT-6.2 Sol or Muse Ember: which to pick
Pick GPT-6.2 Sol for coding: it scores 83.1 on the Vibe Coding Index against 64.9 for Muse Ember.
Pick Muse Ember for output tokens 20.0x cheaper ($0.50 vs $10.00 per million) and work measured by Tax Agent Bench (Vals AI) or Legal Research Bench (Vals AI), where it leads.
- Shared results
- 20
- GPT-6.2 Sol only
- 36
- Muse Ember only
- 3
- Like-for-like categories
- 7 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
GPT-6.2 Sol
GPT-6.2 Sol scores 83.1 against 64.9.
4 and 3 benchmarks behind the scores
Intelligence
AA Intelligence Index
GPT-6.2 Sol
GPT-6.2 Sol leads, 57.3 against 36.0.
Same board, both measured
Web design
Design Arena: websites
No call
Only Muse Ember has a Design Arena: websites result, 1325.
Directional only
General
Vals Index
GPT-6.2 Sol
GPT-6.2 Sol leads, 66.1% against 60.5%.
Same board, both measured
Long documents
Listed context window
GPT-6.2 Sol
GPT-6.2 Sol takes 1.05M tokens against 1.05M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Muse Ember
GPT-6.2 Sol $0.0070, Muse Ember $0.0003 per turn.
List prices
At a glance
- Coding index
- GPT-6.2 Sol 83.1 from 4 benchmarks, Muse Ember 64.9 from 3 benchmarks.
- Widest category
- Reasoning, 31.9 points apart on average over 2 shared benchmarks.
- Widest single gap
- HLE-Diamond (Artificial Analysis run): GPT-6.2 Sol 64.0% against Muse Ember 25.1%, 38.9 points apart on a 0 to 100 scale.
- Output price
- GPT-6.2 Sol $10.00 against Muse Ember $0.50 per million output tokens, 20.0x apart.
- Context window
- GPT-6.2 Sol takes 1.05M tokens, Muse Ember 1.05M.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | GPT-6.2 Sol | Bars | Muse Ember | Ties |
|---|---|---|---|---|
| Coding | 8 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 3 | ||
| Reasoning | 2 | 0 | ||
| Knowledge | 1 | 0 | ||
| Web design | 3 | 0 | ||
| General | 1 | 0 | ||
| All shared boards | 17 | 3 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index80
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| LiveBench coding | 83.7% | Not measured | Only GPT-6.2 Sol measured |
| LMArena WebDev | 1754 | 1608 | GPT-6.2 Sol |
| Terminal-Bench 4 (Vals AI) | 60.6% | 39.4% | GPT-6.2 Sol |
| Vibe Code Bench (Vals AI) | 87.6% | 83.7% | GPT-6.2 Sol |
| Terminal-Bench 4.0 (Artificial Analysis run) | 60.1% | 40.9% | GPT-6.2 Sol |
| FrontierCode 2.0 | 44.4% | 27.2% | GPT-6.2 Sol |
| LMArena Text: coding | 1440 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: software and IT | 1434 | Not measured | Only GPT-6.2 Sol measured |
| LMArena WebDev: React | 1761 | 1616 | GPT-6.2 Sol |
| LMArena WebDev: HTML | 1840 | 1624 | GPT-6.2 Sol |
| GameBench 2 | 94.4 | 58.7 | GPT-6.2 Sol |
| Neither model has a result on the other 41 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| AA Intelligence Index | 57.3 | 36.0 | GPT-6.2 Sol |
| Neither model has a result on the other 1 board here | |||
Agentic workLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| LiveBench research | 90.1% | Not measured | Only GPT-6.2 Sol measured |
| LiveBench browser | 88.2% | Not measured | Only GPT-6.2 Sol measured |
| LiveBench long agentic | 63.6% | Not measured | Only GPT-6.2 Sol measured |
| Tax Agent Bench (Vals AI) | 24.8% | 38.2% | Muse Ember |
| Legal Research Bench (Vals AI) | 37.0% | 49.0% | Muse Ember |
| Harvey's Legal Agent Benchmark (Vals AI) | 12.1% | 25.0% | Muse Ember |
| BioMysteryBench (Vals AI) | 77.8% | Not measured | Only GPT-6.2 Sol measured |
| Terminal-Bench-Science (Vals AI) | 58.6% | Not measured | Only GPT-6.2 Sol measured |
| Vals Science Index | 57.8% | Not measured | Only GPT-6.2 Sol measured |
| APEX-Agents | 66.6% | 64.3% | GPT-6.2 Sol |
| LMArena Search | 1179 | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 33 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| LiveBench reasoning | 90.6% | Not measured | Only GPT-6.2 Sol measured |
| MysteryMechanism (Vals AI) | 36.9% | Not measured | Only GPT-6.2 Sol measured |
| HLE-Diamond (Artificial Analysis run) | 64.0% | 25.1% | GPT-6.2 Sol |
| CritPt v2, hard subset (Artificial Analysis run) | 64.0% | 39.0% | GPT-6.2 Sol |
| ARC-AGI | 97.5% | Not measured | Only GPT-6.2 Sol measured |
| ARC-AGI-2 | 85.8% | Not measured | Only GPT-6.2 Sol measured |
| SimpleBench | 69.9% | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: hard prompts | 1419 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: hard prompts in English | 1413 | Not measured | Only GPT-6.2 Sol measured |
| Kagi LLM Benchmark | 67.0% | Not measured | Only GPT-6.2 Sol measured |
| Humanity's Last Exam (Scale run) | 56.5% | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 15 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| MathArena | 91.2% | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 13 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| AA-Omniscience | 40.4 | 7.7 | GPT-6.2 Sol |
| LMArena Text: science | 1432 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: business and finance | 1383 | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 18 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index30
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| Design Arena: websites | Not measured | 1325 | Only Muse Ember measured |
| Design Arena: UI components | Not measured | 1305 | Only Muse Ember measured |
| Design Arena: game dev | 1437 | 1315 | GPT-6.2 Sol |
| Design Arena: data viz | Not measured | 1314 | Only Muse Ember measured |
| Design Arena: 3D | 1482 | 1357 | GPT-6.2 Sol |
| Design Arena: all code categories | 1405 | 1340 | GPT-6.2 Sol |
| LiveBench design | 80.0% | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| Vals Index | 66.1% | 60.5% | GPT-6.2 Sol |
| LiveBench data analysis | 81.1% | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text | 1414 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: instruction following | 1434 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: creative writing | 1403 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: English | 1419 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: non-English | 1405 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: Chinese | 1434 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: Russian | 1401 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: multi-turn | 1420 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: longer queries | 1427 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: writing and language | 1420 | Not measured | Only GPT-6.2 Sol measured |
| LMArena Text: entertainment and media | 1408 | Not measured | Only GPT-6.2 Sol measured |
| MultiChallenge | 72.0% | Not measured | Only GPT-6.2 Sol measured |
| Neither model has a result on the other 35 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | GPT-6.2 Sol | Muse Ember | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Muse Ember costs less
Repository review
50K input + 3K output tokens
Muse Ember costs less
Agent loop
200K input + 10K output tokens
Muse Ember costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | OpenAI GPT-6.2 Sol | Meta Muse Ember |
|---|---|---|
| Maker | OpenAI | Meta |
| Weights | No public weights listed | No public weights listed |
| Released | November 17, 2026 | November 9, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $2.00 | $0.10 |
| Output $/Mtok | $10.00 | $0.50 |
| Coding index | 83.1 | 64.9 |
| Independent measurements behind it | 4 | 3 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is GPT-6.2 Sol better than Muse Ember for coding?
Pick GPT-6.2 Sol for coding: it scores 83.1 on the Vibe Coding Index against 64.9 for Muse Ember. Pick Muse Ember for output tokens 20.0x cheaper ($0.50 vs $10.00 per million) and work measured by Tax Agent Bench (Vals AI) or Legal Research Bench (Vals AI), where it leads.
Which is cheaper, GPT-6.2 Sol or Muse Ember?
GPT-6.2 Sol costs $2.00 per million input tokens and $10.00 per million output tokens; Muse Ember costs $0.10 and $0.50. Cost per solved task is known for neither of them.
Which has the bigger context window, GPT-6.2 Sol or Muse Ember?
GPT-6.2 Sol, with 1.05M tokens against 1.05M.
How many benchmarks is the GPT-6.2 Sol vs Muse Ember comparison based on?
20 benchmarks have a result for both models as of November 22, 2026. GPT-6.2 Sol's coding index rests on 4 independent benchmarks and Muse Ember's on 3. Where only one model has a result, the page says so instead of naming a winner.