ZMIME
Comparison · 3 models · Updated Nov 18, 2026

GPT-5.6 Sol vs GPT-6.2 Astra vs Muse Wonder

GPT-6.2 Astra comes out ahead, 91 to 88 and 84 on our weighted score, though GPT-5.6 Sol is 2× cheaper per token.

Price in rating: Excluded · Include price

  1. OpenAI

    GPT-5.6 Sol

    Released Jul 9, 2026

    84/100
    • ECI163.1
    • Price$4.00 / $20.00
    • Context1.05M
  2. Our pick

    OpenAI

    GPT-6.2 Astra

    Released Nov 17, 2026New

    91/100
    • ECI174.2
    • Price$8.00 / $40.00
    • Context1.05M
  3. Meta

    Muse Wonder

    Released Oct 12, 2026New

    88/100
    • ECI161.7
    • Price$4.00 / $20.00
    • Context1.05M
01 — Verdict

GPT-6.2 Astra is our pick

GPT-6.2 Astra is the better all-round choice, scoring 91/100 against Muse Wonder (88) and GPT-5.6 Sol (84). It leads on capability. Muse Wonder wins on inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.

  • CapabilityGPT-6.2 AstraCapabilities Index (ECI): GPT-6.2 Astra 174.2 · GPT-5.6 Sol 163.1 · Muse Wonder 161.7
  • Lowest priceGPT-5.6 Sol and Muse WonderGPT-5.6 Sol $8.00 · Muse Wonder $8.00 · GPT-6.2 Astra $16.00 per 1M tokens (3:1 blend)
  • Longest contextAbout the sameGPT-5.6 Sol 1,050,000 · GPT-6.2 Astra 1,050,000 · Muse Wonder 1,048,576 tokens
  • Widest inputsMuse WonderGPT-5.6 Sol: Text, Images, PDFs · GPT-6.2 Astra: Text, Images, PDFs · Muse Wonder: Text, Images, PDFs, Audio, Video
  • Self-hostingNo confirmed open weightsAll three are available only through APIs
How the score is built
MeasureWeightGPT-5.6 SolGPT-6.2 AstraMuse Wonder
CapabilityCapabilities Index (ECI)67%9010088
Inputs & features20%8080100
Context window13%656565
Overall100%84/10091/10088/100
02 — Side by side

Every spec in one table

Highlighted cells lead their row. Dashes mean the data is not published.

GPT-5.6 Sol vs GPT-6.2 Astra vs Muse Wonder specifications side by side
SpecificationGPT-5.6 SolOpenAIGPT-6.2 AstraOpenAIMuse WonderMeta
Capability
Capabilities Index (ECI)163.1174.2 (best)161.7
ECI rank#14 of 158#1 of 158 (best)#16 of 158
GPQA DiamondGraduate-level science questions93.5%96.8% (best)—
FrontierMath Tiers 1–3Research-level mathematics89.1%92.6% (best)82.1%
OTIS Mock AIME 2024–2025Competition mathematics100% (best)100% (best)99.4%
SimpleQA VerifiedShort factual questions69.7%78.5% (best)—
Price per million tokens
Input$4.00 (best)$8.00$4.00 (best)
Output$20.00 (best)$40.00$20.00 (best)
Cached input$0.40$0.40$0.20 (best)
Blended (3:1)$8.00 (best)$16.00$8.00 (best)
Long-context rateOver 272K: $8.00 / $30.00Over 272K: $16.00 / $60.00Same rate
Price sourceOfficial OpenAI APIOfficial OpenAI APIOfficial Meta API
Limits
Context window1,050,000 tokens (best)1,050,000 tokens (best)1,048,576 tokens
Max output128,000 tokens128,000 tokens131,072 tokens (best)
Inputs and features
TextYesYesYes
ImagesYesYesYes
PDFsYesYesYes
AudioNoNoYes
VideoNoNoYes
ReasoningYeslow · medium · high · xhigh · maxYeslow · medium · high · xhigh · maxYesminimal · low · medium · high · xhigh · max
Tool callingYesYesYes
Structured outputYesYesYes
Availability
WeightsProprietaryProprietaryProprietary
API model IDgpt-5.6-solgpt-6.2-astramuse-wonder-1.0
API providers61 (best)1210
ReleasedJul 9, 2026Nov 17, 2026Oct 12, 2026
Knowledge cutoffFeb 16, 2026May 18, 2026—
03 — Cost

What would a month cost?

Enter your expected volume in millions of tokens. List prices only; caching and batch discounts would lower these.

  • GPT-5.6 Sol$80.00
  • GPT-6.2 Astra$160.00
  • Muse Wonder$80.00
04 — Questions

Which should you choose?

Which is better: GPT-5.6 Sol, GPT-6.2 Astra or Muse Wonder?

GPT-6.2 Astra is the better all-round choice, scoring 91/100 against Muse Wonder (88) and GPT-5.6 Sol (84). It leads on capability. Muse Wonder wins on inputs & features. The score weighs capability 67%, inputs & features 20%, context window 13%.

Which is cheaper, GPT-5.6 Sol, GPT-6.2 Astra or Muse Wonder?

GPT-5.6 Sol is cheaper at $4.00 input / $20.00 output per million tokens (official OpenAI API price). Muse Wonder costs $4.00 input / $20.00 output per million tokens (official Meta API price); GPT-6.2 Astra costs $8.00 input / $40.00 output per million tokens (official OpenAI API price). At a typical mix of three input tokens to one output token, that is $8.00 per million tokens for GPT-5.6 Sol versus $8.00 for Muse Wonder (1× as much) and $16.00 for GPT-6.2 Astra (2× as much).

Which scores higher on benchmarks?

GPT-6.2 Astra scores higher on the Capabilities Index (ECI): GPT-6.2 Astra 174.2 (#1 of 158), GPT-5.6 Sol 163.1 (#14 of 158) and Muse Wonder 161.7 (#16 of 158). Their confidence ranges do not overlap (170.3–178.6 vs 160.8–165.6), so the gap is a real one. On individual benchmarks: FrontierMath Tiers 1–3 — GPT-6.2 Astra 92.6%, GPT-5.6 Sol 89.1%, Muse Wonder 82.1%; OTIS Mock AIME 2024–2025 — GPT-5.6 Sol 100%, GPT-6.2 Astra 100%, Muse Wonder 99.4%.

Which is better for coding?

There are no published SWE-bench Verified results for GPT-5.6 Sol, GPT-6.2 Astra and Muse Wonder yet, so there is no like-for-like coding score. On overall capability, GPT-6.2 Astra leads, which tends to carry over to coding, but test on your own codebase. All three support tool calling for agent workflows.

Which has the bigger context window?

Their context windows are effectively the same size: GPT-5.6 Sol 1,050,000, GPT-6.2 Astra 1,050,000 and Muse Wonder 1,048,576 tokens. Maximum output per response: GPT-5.6 Sol up to 128,000, GPT-6.2 Astra up to 128,000, Muse Wonder up to 131,072 tokens.

Which can read images, PDFs, audio or video?

GPT-5.6 Sol accepts text, images and PDFs; GPT-6.2 Astra accepts text, images and PDFs; Muse Wonder accepts text, images, PDFs, audio and video. Muse Wonder handles the widest range of inputs.

Are any of these open source?

No. GPT-5.6 Sol, GPT-6.2 Astra and Muse Wonder are proprietary and only available through APIs and apps.

Which is newer?

GPT-6.2 Astra is the newest, released Nov 17, 2026. Muse Wonder came out Oct 12, 2026; GPT-5.6 Sol came out Jul 9, 2026. Knowledge cutoff: GPT-5.6 Sol Feb 16, 2026, GPT-6.2 Astra May 18, 2026.

How do you decide the winner?

This rating uses capability (67%, independent benchmark results); inputs & features (20%, image, PDF, audio and video input, tool calling, structured output and reasoning); context window (13%, maximum tokens per request, log scale). Dimensions missing for any model are dropped and the remaining weights rescaled, so every model is judged on the same evidence. Specs and prices come from public model listings and the labs’ own API pages; capability scores come from independent benchmark runs. Data updated Nov 18, 2026.