Model comparison
Claude Sonnet 5 vs GPT-6 Sol: same prices, different caches
Claude Sonnet 5 and GPT-6 Sol list the same $2 input and $10 output rates. Cache writes decide the gap: $2.40 against $2.10 for one coding session.
· Prices as of September 28, 2026
Claude Sonnet 5
Anthropic · Released June 30, 2026
Anthropic's balance of speed and intelligence, and a drop-in upgrade for Claude Sonnet 4.6. Anthropic pitches it as close to Claude Opus 4.8 at a lower price.
Claude Sonnet 5 facts and comparisonsGPT-6 Sol
OpenAI · Released September 22, 2026
The mid-priced GPT-6 model, which OpenAI pitches for complex coding and agent workflows and which the Codex docs recommend for complex coding.
GPT-6 Sol facts and comparisons
The short answer
Claude Sonnet 5 and GPT-6 Sol list identical per-token prices, $2 input and $10 output per million, so uncached work costs the same on both. On the example agentic coding session GPT-6 Sol comes to $2.10 against $2.40, a $0.30 gap that comes entirely from Anthropic's pricier 1-hour cache writes. The practical choice is the tool you work in: Sonnet 5 lives in Claude Code, and GPT-6 Sol is the model the Codex docs recommend for complex coding.
Choose Claude Sonnet 5 if
- Claude Code is where you work, and its sonnet alias maps to Claude Sonnet 5 on the Anthropic API.
- You want Anthropic's agentic pitch: planning and using tools like browsers and terminals on its own.
- You are moving off Claude Sonnet 4.6, since Anthropic describes Sonnet 5 as a drop-in upgrade.
- You want a price that has already settled: the launch price became the standard price on August 10, 2026, and a planned increase was cancelled.
Choose GPT-6 Sol if
- You work in Codex, whose docs recommend GPT-6 Sol for complex coding and suggest moving to it from GPT-5.6 Sol and GPT-5.6 Terra.
- Your sessions write a lot to the cache: OpenAI bills every write at 1.25x input, $2.50 per million, with no 2x tier.
- You want each cached prefix to stay reusable for at least 30 minutes after its last use.
- You prefer a medium reasoning default, which is where GPT-6 Sol starts in both the API and Codex.
Side by side
Specs and prices
| Fact | Claude Sonnet 5 | GPT-6 Sol |
|---|---|---|
| Maker | Anthropic | OpenAI |
| API model id | claude-sonnet-5 | gpt-6-sol |
| Released | June 30, 2026 | September 22, 2026 |
| Status | Current | Current |
| Context window | 1M tokens | 1.05M tokens |
| Max output | 128K tokens | 128K tokens |
| Open weights | No | No |
| Input, per 1M tokens | $2 | $2 |
| Cache hit, per 1M | $0.20 | $0.20 |
| Cache write, per 1M | $2.50 (5-minute), $4 (1-hour) | $2.50 |
| Output, per 1M tokens | $10 | $10 |
| Runs in | Claude Code, Cursor, OpenCode, OpenRouter, and GitHub Copilot | Codex, OpenCode, OpenRouter, and GitHub Copilot |
Standard API rates in US dollars, as published by each maker (Claude Sonnet 5: September 26, 2026; GPT-6 Sol: September 28, 2026). Batch and priority tiers, taxes, and subscription plans are not included. Claude Sonnet 5: The full 1M context window is billed at standard rates. The launch price became the standard price on August 10, 2026, and a planned increase was cancelled. GPT-6 Sol: Requests over 272K input tokens cost 2x for input and cache and 1.5x for output, for the whole request.
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 Sonnet 5 | GPT-6 Sol |
|---|---|---|
| Agentic coding session, 100K input, 400K written to cache, 2M read from cache, 50K output | $2.40 | $2.10 |
| Large one-off review, 150K input with no cache hits, 10K output | $0.40 | $0.40 |
| Output-heavy generation, 30K input, 80K output | $0.86 | $0.86 |
| A month of sessions, 110 sessions: 5 a day, 22 working days | $264.00 | $231.00 |
| Where the session’s cost goes | ||
| Cache writes | $1.30 | $1.00 |
| Cache reads | $0.40 | $0.40 |
| Uncached input | $0.20 | $0.20 |
| Output | $0.50 | $0.50 |
- caching saves on the session with Claude Sonnet 5 (56%)
- $3.10
- caching saves on the session with GPT-6 Sol (62%)
- $3.40
Why GPT-6 Sol costs less on a session when the rates match
Line up the rate cards and they read the same. Claude Sonnet 5 and GPT-6 Sol both charge $2 per million input tokens, $0.20 per million cache hits, $10 per million output tokens, and $2.50 per million for a standard cache write. The large one-off review costs $0.40 on both, and the output-heavy generation costs $0.86 on both.
The difference sits in one row. Anthropic sells two cache lifetimes: a 5-minute write at 1.25x input and a 1-hour write at 2x, which is $4 per million on Sonnet 5. OpenAI has a single write price. The example session writes 400K tokens and, on Claude, splits them evenly between the two lifetimes, so writes cost $1.30 on Sonnet 5 against $1.00 on GPT-6 Sol. Reads ($0.40), fresh input ($0.20), and output ($0.50) are identical, which leaves a $0.30 gap and a 13% saving for GPT-6 Sol.
Over 110 sessions a month that becomes $264.00 against $231.00, a $33.00 difference. It also depends on how Claude Code is set up. With an API key the main conversation uses the 5-minute cache, which costs exactly what GPT-6 Sol charges for a write, so the two would price this session the same. On a Claude subscription it uses the 1-hour cache, but subscription plans are priced differently from API rates anyway.
How Anthropic and OpenAI pitch these two models
Anthropic calls Sonnet 5 its balance of speed and intelligence and says it is "built to be the most agentic Sonnet model yet." It claims performance close to Claude Opus 4.8 at lower prices, and better reasoning, tool use, and coding than Claude Sonnet 4.6. Adaptive thinking is on by default and the default effort is high.
OpenAI describes GPT-6 Sol as "built to power complex coding and agentic workflows," and it is the mid-priced model of the GPT-6 family, between Luna and Astra. Its reasoning effort defaults to medium. Released on September 22, 2026, it is the newer of the two by almost three months.
Those defaults matter for cost. Output is $0.50 of each session here, and a model thinking at high effort can write more tokens than one at medium. The tokenizers also differ, so the same source file does not count as the same number of tokens on both. Identical rates do not promise identical costs on your own work.
Context limits and the 272K line
GPT-6 Sol has a 1.05M context window, of which up to 922K can be input, and Sonnet 5 has 1M. Each can write up to 128K tokens of output. For most coding sessions neither limit comes into play.
GPT-6 Sol does carry a long-context rule: requests over 272K input tokens cost 2x for input and cache and 1.5x for output, and the higher rate applies to the whole request, not just the tokens past the line. The example session keeps each request under 200K, so no tier applies to it. An agent that loads a whole repository or a long log into one request can cross it.
To see how your own sessions split across the two, EveryToken reads Claude Code and Codex history on your Mac and prices each request at API rates, including what the cache saved or cost. Treat the result as an API-equivalent estimate at published rates.
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 Sonnet 5 and GPT-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-6 Sol cheaper than Claude Sonnet 5?
Per token, no: input, output, cache hits, and 5-minute cache writes are priced the same. On the example cached session GPT-6 Sol costs $2.10 against $2.40, because Anthropic's 1-hour writes cost 2x input. With no cache involved, the two cost exactly the same.
What happens above 272K input tokens on GPT-6 Sol?
The whole request is billed at 2x for input and cache and 1.5x for output. GPT-6 Sol accepts up to 922K input tokens, so large prompts are possible, just at the higher rate.
Which coding tools offer each model?
Claude Sonnet 5 runs in Claude Code, Cursor, OpenRouter, OpenCode, and GitHub Copilot. GPT-6 Sol runs in Codex, OpenRouter, OpenCode, and GitHub Copilot.
Which model should I switch to from GPT-5.6 Sol?
The Codex docs suggest GPT-6 Sol as the move from GPT-5.6 Sol, GPT-5.6 Terra, and GPT-5.4. If you are open to changing tools as well, Claude Sonnet 5 lists the same rates, so the decision comes down to workflow and caching rather than price.
Sources
- Anthropic: Pricing
- Anthropic docs: Claude Sonnet 5
- Anthropic: Introducing Claude Sonnet 5
- Claude Code docs: Model configuration
- Cursor docs: Claude Sonnet 5
- OpenRouter: Claude Sonnet 5
- OpenCode docs: Zen
- GitHub Docs: Supported AI models in Copilot
- OpenAI: API pricing
- OpenAI docs: GPT-6 Sol
- OpenAI: Introducing GPT-6 Sol and Luna
- OpenAI: API changelog
- Codex docs: Models
- OpenRouter: GPT-6 Sol
- Anthropic: Prompt caching
- OpenAI: Prompt caching