Claude Opus 5.5 vs GPT-6.1 Sol
Pricing, benchmarks, and use case comparison
Verdict
Pick Claude Opus 5.5 for the broadest published coding and computer-use results and flat pricing across its 1M context; pick GPT-6.1 Sol for near-even agentic coding at half the per-token price ($2/$10 vs $4/$20).
Specs comparison
| Claude Opus 5.5 | GPT-6.1 Sol | |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Type | Closed source | Closed source |
| Context window | ✓1M | 1,050,000 |
| Input / 1M tokens | $4.00 | ✓$2.00 |
| Output / 1M tokens | $20.00 | $10.00 |
| Release date | 2026-09 | 2026-09 |
Benchmarks
| Benchmark | Claude Opus 5.5 | GPT-6.1 Sol |
|---|---|---|
| Terminal-Bench 4.0 | 66.4% (xhigh effort) | - |
| Anthropic Automated Behavioral Audit | Strongest-performing model tested | - |
| SWE-bench Pro | 89.9% (max effort) | - |
| DeepSWE v1.1 | 74.2% (max effort) | ✓75.2% (high effort) |
| OSWorld 2.1 | 81.8% (max effort) | - |
| OSWorld 2.0 offline subset | - | 71.4% (max effort) |
| Terminal-Bench Science 0.1 (Max Effort) | - | Cost: $5.47/task |
| AutomationBench (Medium Effort) | - | 2.2 points above Claude Opus 5.5 |
| Factual Accuracy (Low Reasoning Effort) | - | 7.7% error rate |
Scores sourced from official provider release posts and independent benchmark aggregators.
Capability and benchmarks
The one coding benchmark both vendors report is DeepSWE v1.1, and it is close: Claude Opus 5.5 scores 74.2% at max effort, while GPT-6.1 Sol scores 75.2% at high effort and 71.9% at max. Beyond that, the published results do not overlap. Anthropic reports 89.9% SWE-bench Pro, 66.4% Terminal-Bench 4.0 (xhigh effort) and 81.8% OSWorld 2.1 for Opus 5.5. OpenAI reports 71.4% on the OSWorld 2.0 offline subset for Sol, which is a different test set, and says Sol scores 2.2 points above Opus 5.5 on AutomationBench at medium effort. OpenAI positions Sol as close to its GPT-6 Astra flagship on agentic coding, computer use and professional work. All of these are vendor-run numbers, with each lab choosing its own harness and effort setting.
Price and context
GPT-6.1 Sol costs $2 input / $10 output per 1M tokens (cached input $0.10), half of Opus 5.5's $4/$20 (cache hits $0.20). OpenAI's Terminal-Bench Science run puts Sol at $5.47 per task against $23.21 for Opus 5.5 at max effort. The gap narrows on long prompts: Sol has a 1,050,000-token context (922K max input), but any request above 272K input tokens pays double for input and 1.5x for output, while Opus 5.5 bills its full 1M context at the standard rate. Both cap output at 128K tokens and accept text and image input.
- Opus 5.5: $4/$20, 1M context, no long-context surcharge, Anthropic's recommended starting model.
- GPT-6.1 Sol: $2/$10 up to 272K input tokens, then $4/$15 for the whole request.
Which to pick
- Pick Claude Opus 5.5 for long-running agentic coding, computer use and knowledge work where its SWE-bench Pro and OSWorld results matter, and for long-context jobs that would cross Sol's 272K surcharge line.
- Pick GPT-6.1 Sol for repeated, cost-sensitive agent workloads, where half-price tokens and a lower cost per task add up, and if your stack already runs on the OpenAI API or Codex.
Which should you choose?
Choose Claude Opus 5.5 if...
- →When you need Fable-level performance at mid-tier pricing
- →For long-context coding and knowledge work
- →When cost-efficiency matters alongside capability
Choose GPT-6.1 Sol if...
- →When you need Astra-level intelligence but have a 1/5 budget constraint
- →For repeated and long-running workloads where cost per task matters
- →When coding performance and computer use are primary requirements