AI API Comparator

Pick up to 3 models · pricing, context and capabilities · simulate your product cost

Data updated on September 30, 2026· 217 models

AnthropicSonnet 5

Sonnet 5 is Anthropic's most capable Sonnet-class model, with frontier performance across coding, agents, and professional work.

View sheet →

Reasoning estimated
Speed estimated
Input
Output
Context
1 M tokens
Max output
128 K
Input, $ per 1M tokens
$2.00
Output, $ per 1M tokens
$10.00
Cache read, $ per 1M
$0.200
Batch (input / output)
$1.00 / $5.00
Tool calling
Yes
Released
06/30/2026
Knowledge cutoff
2026-01-31
API id
anthropic/claude-sonnet-5
OpenAIGPT-6 Sol

GPT-6 Sol is the cost-efficient high-end model in OpenAI's GPT-6 series, positioned below the flagship GPT-6 Astra and above the fast GPT-6 Luna tier.

View sheet →

Reasoning estimated
Speed estimated
Input
Output
Context
▲1.1 M tokens
Max output
128 K
Input, $ per 1M tokens
$2.00
Output, $ per 1M tokens
$10.00
Cache read, $ per 1M
$0.200
Batch (input / output)
$1.00 / $5.00
Tool calling
Yes
Released
▲09/22/2026
Knowledge cutoff
2026-04-20
API id
openai/gpt-6-sol
Third model (optional)
text image audio video PDF / file reasoning (1 to 5, estimated by family) speed (1 to 4, estimated by family)▲ best value among the selected

Verdict generated from the data

  • GPT-6 Sol: largest context (1.1 M)
  • GPT-6 Sol: the newest (09/22/2026)

Claude Sonnet 5 and GPT-6 Sol cost the same ($2.00 input, $10.00 output per million tokens). GPT-6 Sol accepts more context (1.1 M). Claude Sonnet 5 and GPT-6 Sol offer a batch tier at half price for jobs that do not need an immediate answer.

What would your product cost?

Estimate the monthly cost with your app's real volume. List prices, no cache or batch discounts.

Claude Sonnet 5
$70.00/mo
GPT-6 Sol
$70.00/mo

They cost the same at this volume.

The 6 cheapest for this scenario (chat, last 12 months)

  1. 1Lyria 3 Pro Preview Google$0.00
  2. 2Lyria 3 Clip Preview Google$0.00
  3. 3Qwen3.7 Flash Qwen$0.97
  4. 4DeepSeek V4 Flash 0731 DeepSeek$1.55
  5. 5Ministral 3 3B 2512 Mistral$1.90
  6. 6Qwen3.5-Flash Qwen$2.02

Paste your prompt: tokens and cost per model

Rough estimate (about 4 characters per token). Nothing is sent to any server.

76tokens approx.287 characters

Cost of sending it 1,000 times (input only)

  1. 1Claude Sonnet 5 Anthropic$0.152
  2. 2GPT-6 Sol OpenAI$0.152
  3. 3Claude Sonnet 5.5 Anthropic$0.152
  4. 4Claude Opus 5.5 Anthropic$0.304
  5. 5Claude Opus 5 Anthropic$0.380
  6. 6Claude Opus 4.8 Anthropic$0.380
  7. 7Claude Fable 5.1 Anthropic$0.760
  8. 8Claude Fable 5 Anthropic$0.760

Popular comparisons

Pricing and spec sheets for the models developers use through an API: Claude (Sonnet, Haiku, Opus, Fable), GPT, Gemini, DeepSeek, Grok, Llama, Mistral, Qwen and Kimi. Data comes from the providers' public catalogs and syncs once a day, so what you see is what they charge today.

  • 100% free
  • Updated daily
  • No signup
  • Official sources

Frequently asked questions

Where do the prices come from and how often are they updated?

From the public catalogs of OpenRouter, models.dev and LiteLLM, which publish each provider's list prices. They sync once a day and the last update date is shown above the comparison.

What does price per million tokens mean?

It is what the provider charges per million tokens you send (input) or the model generates (output). A token is roughly 4 characters. Output almost always costs more than input.

How do I calculate what my app would cost?

Use the simulator: enter how many requests you make per month and how many input and output tokens each one has. The cost uses list prices with no cache or batch discounts, so it is a reasonable ceiling.

What are the batch tier and cache read?

Batch is a queue for jobs without an immediate answer (hours) at half price. Cache read is the reduced price when the provider already has a repeated part of the prompt in memory, such as a long system prompt.

Are the reasoning and speed scores measured?

Not yet, they are estimates by family: Opus, Pro or Astra models with a reasoning mode score high on reasoning; Flash, Mini or Haiku score high on speed. That is why they carry the "estimated" label. They will be replaced by measurements once the sync has them.

Does it include ChatGPT Plus, Claude Max or Gemini Advanced?

No. This tool compares models used through an API to build products, not the subscription plans of the chat apps.

How much does the Claude or GPT API cost?

It depends on the model: pick the one you care about in any column and you will see input and output price per million tokens, context and capabilities, with the data update date.

Sources: OpenRouter, models.dev and LiteLLM. Prices in USD per million tokens, list price, no volume discounts. Prompt tokens are an estimate; each provider's tokenizer may differ.