Model comparison
Claude Opus 5 vs GPT-5.6 Sol: coding costs compared
Claude Opus 5 lists 25% above GPT-5.6 Sol, and Anthropic's 1-hour cache writes stretch that to 43% on a coding session. Promotions, long context, successors.
· Prices as of September 28, 2026
Claude Opus 5
Anthropic · Released July 24, 2026 · Previous generation
The previous everyday Opus, pitched as close to Claude Fable 5 at half the price. Anthropic now recommends moving to Claude Opus 5.5.
Claude Opus 5 facts and comparisonsGPT-5.6 Sol
OpenAI · Released July 9, 2026 · Previous generation
The flagship of the July 2026 GPT-5.6 family, now succeeded by GPT-6 Sol and on promotional pricing.
GPT-5.6 Sol facts and comparisons
The short answer
GPT-5.6 Sol lists 20% below Claude Opus 5 on every rate, and the example agentic coding session costs $4.20 on it against $6.00, because OpenAI's cache writes cost 1.25x input while Anthropic's 1-hour writes cost 2x. Both are previous-generation models: Anthropic recommends Claude Opus 5.5 over Opus 5, and Codex suggests GPT-6 Sol over GPT-5.6 Sol, whose current rates are promotional.
Choose Claude Opus 5 if
- You work in Claude Code, where Opus 5 remains selectable as a legacy model.
- Your prompts go past 272K input tokens, where GPT-5.6 Sol's rates rise and Opus 5 bills its full 1M window at standard rates.
- You want a 1-hour cache for sessions with long pauses between turns, which Anthropic offers at 2x input.
Choose GPT-5.6 Sol if
- You use Codex cloud chats on a ChatGPT plan, which run on GPT-5.6 Sol.
- You want cheaper cache writes: OpenAI bills a write at 1.25x input, $5 per million.
- Your requests stay under 272K input tokens and you can use the promotional rates, available at least through November 21, 2026.
- OpenAI's pitch of token efficiency and stronger frontend aesthetics fits your work.
Side by side
Specs and prices
| Fact | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| Maker | Anthropic | OpenAI |
| API model id | claude-opus-5 | gpt-5.6-sol |
| Released | July 24, 2026 | July 9, 2026 |
| Status | Previous generation | Previous generation |
| Context window | 1M tokens | 1.05M tokens |
| Max output | 128K tokens | 128K tokens |
| Open weights | No | No |
| Input, per 1M tokens | $5 | $4 |
| Cache hit, per 1M | $0.50 | $0.40 |
| Cache write, per 1M | $6.25 (5-minute), $10 (1-hour) | $5 |
| Output, per 1M tokens | $25 | $20 |
| Runs in | Claude Code, Cursor, OpenCode, OpenRouter, and GitHub Copilot | Codex, Cursor, OpenCode, OpenRouter, and GitHub Copilot |
Standard API rates in US dollars, as published by each maker (Claude Opus 5: September 26, 2026; GPT-5.6 Sol: September 28, 2026). Batch and priority tiers, taxes, and subscription plans are not included. Claude Opus 5: The full 1M context window is billed at standard rates. Fast mode, a research preview on the Claude API, costs $10 input and $50 output per million tokens. GPT-5.6 Sol: Requests over 272K input tokens cost 2x for input and cache and 1.5x for output, for the whole request. These are promotional rates, available at least through November 21, 2026.
Cost
What the same work costs
The same token counts, priced at each model’s published rates. Cached tokens are billed at each maker’s cache prices, so the session shows what caching is worth on each model.
Real sessions differ: the two models count the same code as different numbers of tokens, and reasoning settings change how much each one writes. Your own history is the real test.
| Workload | Claude Opus 5 | GPT-5.6 Sol |
|---|---|---|
| Agentic coding session, 100K input, 400K written to cache, 2M read from cache, 50K output | $6.00 | $4.20 |
| Large one-off review, 150K input with no cache hits, 10K output | $1.00 | $0.80 |
| Output-heavy generation, 30K input, 80K output | $2.15 | $1.72 |
| A month of sessions, 110 sessions: 5 a day, 22 working days | $660.00 | $462.00 |
| Where the session’s cost goes | ||
| Cache writes | $3.25 | $2.00 |
| Cache reads | $1.00 | $0.80 |
| Uncached input | $0.50 | $0.40 |
| Output | $1.25 | $1.00 |
- caching saves on the session with Claude Opus 5 (56%)
- $7.75
- caching saves on the session with GPT-5.6 Sol (62%)
- $6.80
Where the 43% session gap comes from
At list prices, Claude Opus 5 costs 25% more than GPT-5.6 Sol on every line: $5 against $4 per million input tokens, $25 against $20 for output, and $0.50 against $0.40 for a cache hit. Uncached work reflects that directly, with the large one-off review at $1.00 against $0.80 and the output-heavy generation at $2.15 against $1.72.
Cache writes are where the makers differ. Anthropic offers two write lifetimes, 5 minutes at 1.25x input and 1 hour at 2x, while OpenAI bills every write at 1.25x from GPT-5.6 on. The example session writes 400K tokens, half of them to Anthropic's 1-hour cache, so writes cost $3.25 on Opus 5 and $2.00 on GPT-5.6 Sol. That $1.25 is most of the $1.80 difference, and it is why the session gap is 43% while the list gap is 25%.
At 110 sessions a month, the projection is $660.00 on Opus 5 and $462.00 on GPT-5.6 Sol at API rates. Caching saves 56% on Opus 5 and 62% on GPT-5.6 Sol against the same tokens sent uncached, a difference that again comes down to what it costs to write the cache.
Cache lifetimes and long prompts work differently
The cheaper OpenAI write doesn't buy the same thing. On GPT-5.6 and later, a cached prefix stays reusable for at least 30 minutes after its last use, caching starts at 1,024 input tokens, and you can mark up to four explicit breakpoints. Anthropic's caches restart their lifetime on every hit at no charge, and in Claude Code the main conversation uses the 1-hour cache on a subscription and the 5-minute cache with an API key.
Long prompts are the other split. GPT-5.6 Sol charges 2x for input and cache and 1.5x for output on any request over 272K input tokens, for the whole request. Opus 5 bills its full 1M context window at standard rates. The example session keeps each request under 200K tokens, so the tables don't show that tier, but an agent that loads a very large codebase into each request would pay it on GPT-5.6 Sol.
The GPT-5.6 Sol rates here are also promotional, available at least through November 21, 2026. Check OpenAI's pricing before planning spend past that date.
Two previous-generation models and their successors
Both models have been superseded. Anthropic pitched Opus 5 as close to the frontier intelligence of Claude Fable 5 at half the price, designed for everyday coding and knowledge work, and it was Claude Code's opus default until Claude Opus 5.5 replaced it. Anthropic now recommends moving to Opus 5.5.
OpenAI describes GPT-5.6 Sol as the "Flagship model for complex professional work" of its July 2026 GPT-5.6 family, and credits it with token efficiency, better frontend aesthetics, and programmatic tool calling, where the model writes JavaScript that calls tools and processes their output. Codex suggests moving to GPT-6 Sol, except in Codex cloud chats on ChatGPT plans, which still use GPT-5.6 Sol. The API id gpt-5.6 points to it.
Where they run differs too. Opus 5 is in Claude Code, Cursor, OpenRouter, OpenCode, and GitHub Copilot, and GPT-5.6 Sol is in Codex, Cursor, OpenRouter, OpenCode, and GitHub Copilot. The two makers tokenize text differently, so the same prompt won't count as the same number of tokens on both, and OpenAI's token-efficiency claim is about output length, which the fixed workloads here can't capture.
Prompt caching
How each maker bills cached tokens
Anthropic
Claude caches a prompt prefix up to a breakpoint. One top-level cache_control field places the breakpoint automatically and moves it as the conversation grows, or you can mark up to 4 blocks yourself. Claude Code manages caching for you.
A 5-minute cache write costs 1.25x the input price and a 1-hour write costs 2x. A cache hit costs 0.1x input on most models, 0.05x on Claude Opus 5.5, and 0.025x on Claude Fable 5.1 and Claude Mythos 5.1. Every hit restarts the cache lifetime at no charge.
Anthropic's rule of thumb: a 5-minute write pays for itself after one cache read, and a 1-hour write after two.
In Claude Code, the main conversation uses the 1-hour cache on a Claude subscription and the 5-minute cache with an API key. Each model has its own cache, so switching models starts over.
Source: Anthropic: Prompt caching
OpenAI
Prompt caching is on by default. From GPT-5.6 on you can also mark up to four explicit cache breakpoints, while GPT-5.5 and earlier cache automatically only.
From GPT-5.6 on, a cache write costs 1.25x the uncached input price and a cache hit costs 0.1x. GPT-5.5 and earlier add no charge for writing the cache: written tokens are billed as ordinary input, and a hit costs 0.1x on the models compared here.
On GPT-5.6 and later, a cached prefix stays reusable for at least 30 minutes after its last use, and caching starts at 1,024 input tokens.
Source: OpenAI: Prompt caching
Your own numbers
See what Claude Opus 5 and GPT-5.6 Sol really cost you.
everyaitoken reads your Claude Code, Cursor, OpenCode, OpenRouter, and Codex history on your Mac and prices every request at API rates, with what caching saved or cost. $9 once.
FAQ
Questions
Is GPT-5.6 Sol cheaper than Claude Opus 5?
At current rates, yes. Every rate is 20% lower, and the example session costs $4.20 against $6.00 because OpenAI's cache writes cost less. GPT-5.6 Sol's rates are promotional, available at least through November 21, 2026.
Why does caching save a larger share on GPT-5.6 Sol?
Writing to its cache costs 1.25x input, while half the session's writes on Opus 5 go to the 1-hour cache at 2x. Caching saves 62% on GPT-5.6 Sol and 56% on Opus 5. Opus 5 still saves more in dollars, $7.75 against $6.80, because its input price is higher.
What should I use instead of these two models?
Anthropic recommends Claude Opus 5.5 over Opus 5, and Codex suggests GPT-6 Sol over GPT-5.6 Sol. Both earlier models stay available.
Can I track Claude Opus 5 and GPT-5.6 Sol costs in one place?
EveryToken reads local history from Claude Code, Codex, Cursor, OpenCode, and OpenRouter on your Mac and prices each request at API rates, so spend on both models shows side by side with what caching saved on each.
Sources
- Anthropic: Pricing
- Anthropic docs: Claude Opus 5
- Anthropic: Introducing Claude Opus 5
- Claude Code docs: Model configuration
- Cursor docs: Claude Opus 5
- OpenRouter: Claude Opus 5
- OpenCode docs: Zen
- GitHub Docs: Supported AI models in Copilot
- OpenAI: API pricing
- OpenAI docs: GPT-5.6 Sol
- OpenAI: Using GPT-5.6
- OpenAI: API changelog
- Codex docs: Models
- Codex docs: Pricing
- Cursor docs: Models and pricing
- OpenRouter: GPT-5.6 Sol
- Anthropic: Prompt caching
- OpenAI: Prompt caching