Download

Claude Opus 5.5 vs GPT-5.6 Sol

At the same $4 input and $20 output per 1M tokens, Claude Opus 5.5 scores higher than GPT-5.6 Sol on both benchmarks they share: 66.4% against 37.3% on Terminal-Bench 4.0 and 54.4% against 47.5% on FrontierCode 1.1. Opus 5.5 also has the cheaper cache, $0.20 against $0.40.

Same price, cheaper cache

Uncached, our sessions cost the same on both: $21.60 for the feature session and $195 for a long agent day. With 90% of prompts read from the cache, Opus 5.5 is cheaper, $5.18 against $6.05 and $41.10 against $49.20, because cached input on GPT-5.6 Sol costs twice as much.

GPT-5.6 Sol reads up to 1.05M tokens and Opus 5.5 up to 1M; both write up to 128K tokens in a reply.

Where the scores come from

Both GPT-5.6 Sol scores reached us through Anthropic's charts: the Terminal-Bench 4.0 figure is OpenAI's own, and Anthropic quoted it alongside the FrontierCode 1.1 one. Opus 5.5 ran Terminal-Bench at xhigh effort. On Humanity's Last Exam with tools Opus 5.5 scored 67.7%, and we have no score for GPT-5.6 Sol.

Dates

GPT-5.6 Sol was first listed on the API on 9 July 2026; Anthropic released Opus 5.5 two months later, on 22 September. OpenAI has released newer Sol models since, and Opus 5.5 against GPT-6.1 Sol compares Opus with the latest of them.

Side by side

Claude Opus 5.5 and GPT-5.6 Sol: specifications and prices
Claude Opus 5.5GPT-5.6 Sol
MakerAnthropicOpenAI
Released22 September 2026[1]9 July 2026[2] first API listing
Context window1M tokens[1]1.05M tokens[3]
Max output128K tokens[1]128K tokens[3]
Input price$4 per 1M tokens[1]$4 per 1M tokens[3]
Output price$20 per 1M tokens[1]$20 per 1M tokens[3]
Cached input$0.20 per 1M tokens[4]$0.40 per 1M tokens[3]
Long promptsOne priceOne price
Price changeNone announcedNone announced
Open weightsNo[1] Offered through the Claude API and cloud platforms; no weights are published.No[5] Listed as an API model, not in OpenAI's open-weight section.
Readstext, image[1]text, image[3]

Benchmarks

Scores as the makers published them, each with its source and date. Makers run these at different settings, so read a gap of a few points as a tie.

Published benchmark scores (higher is better)
Claude Opus 5.5GPT-5.6 Sol
Terminal-Bench 4.0Claude Opus 5.5: 66.4%66.4%GPT-5.6 Sol: 37.3%37.3%FrontierCode 1.1 (Main)Claude Opus 5.5: 54.4%54.4%GPT-5.6 Sol: 47.5%47.5%Humanity's Last Exam (with tools)Claude Opus 5.5: 67.7%67.7%GPT-5.6 Sol: not published
Benchmarks: Claude Opus 5.5 and GPT-5.6 Sol
BenchmarkClaude Opus 5.5GPT-5.6 Sol
Terminal-Bench 4.0Agentic coding tasks in a real terminal.66.4%[6] Run at xhigh effort.37.3%[6] As reported by OpenAI, quoted by Anthropic.
FrontierCode 1.1 (Main)Hard software engineering tasks.54.4%[6] Max effort.47.5%[6] Quoted by Anthropic.
Humanity's Last Exam (with tools)Expert-level questions across many fields, tools allowed.67.7%[6]Not published

What a coding session costs

Our own estimate from the list prices above, for three sessions we define: a coding agent resends the conversation with every call, so prompts are large and grow. "Cached" assumes 90% of prompt tokens are read from the prompt cache and leaves out cache write fees. A long-prompt price applies when a call's prompt is over its threshold. Work out your own mix with the token cost calculator.

Feature session at list prices, US dollars (lower is cheaper)
Claude Opus 5.5GPT-5.6 Sol
No cachingClaude Opus 5.5: $21.60$21.60GPT-5.6 Sol: $21.60$21.6090% of prompt tokens from cacheClaude Opus 5.5: $5.18$5.18GPT-5.6 Sol: $6.05$6.05
Session costs: Claude Opus 5.5 and GPT-5.6 Sol
SessionClaude Opus 5.5Claude Opus 5.5 cachedGPT-5.6 SolGPT-5.6 Sol cached
Quick fix10 calls of 25K prompt and 1.5K output tokens$1.30$0.45$1.30$0.49
Feature session60 calls of 80K prompt and 2K output tokens$21.60$5.18$21.60$6.05
Long agent day300 calls of 150K prompt and 2.5K output tokens$195$41.10$195$49.20

Value for money and release gap

Claude Opus 5.5 is dated 2 months later (75 days; GPT-5.6 Sol: first API listing). Price per point divides our feature-session cost by a published score: lower means more score for the money. It is a rough lens, not a ranking.

Value for money: Claude Opus 5.5 and GPT-5.6 Sol
MeasureClaude Opus 5.5GPT-5.6 Sol
Input tokens per $1250K250K
Filling the whole context window once$4.00$4.20
Feature session per point of Terminal-Bench 4.0$0.33$0.58
Feature session per point of FrontierCode 1.1 (Main)$0.40$0.45

Questions

Do Opus 5.5 and GPT-5.6 Sol cost the same?

Per token, yes: $4 per 1M input tokens and $20 per 1M output tokens on both. Cached input differs: $0.20 on Opus 5.5 and $0.40 on GPT-5.6 Sol.

Which is better at agentic coding?

On Terminal-Bench 4.0, Opus 5.5 scored 66.4% and GPT-5.6 Sol 37.3%, a gap of 29.1 points. The GPT-5.6 Sol score is OpenAI's own, as quoted by Anthropic.

Which has the bigger context window?

GPT-5.6 Sol reads up to 1.05M tokens, Opus 5.5 up to 1M. Both write up to 128K tokens in a reply.

Most compared

Any other two models: pick them in the lists at the top of the page.

More tools

Sources

  1. platform.claude.com: docs/en/models/opus-5-5/overview, read
  2. github.blog: changelog/2026-07-09-openais-gpt-5-6-sol-terra-and-luna-are-now-available-in-github-copilot, read
  3. developers.openai.com: api/docs/models/gpt-5.6-sol, read
  4. platform.claude.com: docs/en/about-claude/pricing, read
  5. developers.openai.com: api/docs/models/all, read
  6. anthropic.com: claude-opus-5-5, read

Numbers from the makers' own pages, checked by hand, last read . List prices in US dollars before tax. Spotted a mistake? office@shuriken.sh.

Made by Shuriken, the harness for coding agents. Its hot swapping moves your chats to another of your accounts or API keys when one nears its limit. Use only accounts you own and are entitled to use.

Try Shuriken on your own computer

Add your Claude, Codex and Gemini accounts once. Shuriken switches between them so your agents keep working.

Download Pricing
Download on the App StoreGet it on Google PlayComing soon