OpenAI
GPT-6.1 Sol
$90.00/ month
$0.009 / request
Free tools / Artificial intelligence
What will your AI project cost?
Turn your usage into an estimate. Compare models by cost and application profile, with transparent terms and sources.
11models to explore4 providers · transparent sources
OpenAI
Claude
Gemini
DeepSeek
Open comparison. No signup.
01 / Define your scenario
Catalog checked on · USD / BRL
How much will you use?
Start with the example: 2,000 input tokens, 500 output tokens and 10,000 requests per month. Adjust the values or paste your own text.
Input includes instructions, history and documents. Output is an assumption: include reasoning tokens even when hidden from the answer. The same task can consume different counts across models.
Advanced settingsUSD · requirements and cache
Filters check support. The estimate covers text; images and tool execution require additional budgeting.
Initial reference FX: ECB, Oct 16, 2026. Adjust for your scenario. Excludes taxes and card exchange rates. Source
Caching and batch processing
Reads and writes partition input without double counting; their sum must be at most 100%. Cache depends on provider prefix, minimum and retention. Batch with cache and explicit storage are not modeled.
02 / Your comparison · 3/4
What is the monthly cost?
The same text volume across selected models. Costs update as you adjust the scenario.
3 of 4 models selected
Anthropic
Claude Sonnet 5.5
$90.00/ month
$0.009 / request
Gemini 3.8 Flash
$33.75/ month
$0.003375 / request
Text API estimate. Excludes taxes, media, tools and infrastructure.
Quality: limited comparison. Selected models have a similar usage profile; this does not establish equivalent performance. This catalog has no validated shared benchmark. Compare using your examples before choosing.
Compare capabilities and terms
How to evaluate quality for this task
Compare factual answers, tone and response time using examples in your language.
Starting profiles to evaluate: Volume and economy / General purpose. Editorial suggestion to validate with your data.
Profiles group provider positioning and intended use. They are not intelligence scores. Without shared tests with controlled versions and settings, quality equivalence remains unverified.
| Criterion | GPT-6.1 SolOpenAI | Claude Sonnet 5.5Anthropic | Gemini 3.8 FlashGoogle |
|---|---|---|---|
| Monthly estimate | $90.00 | $90.00 | $33.75 |
| Usage profile | General purpose | General purpose | General purpose |
| Image / JSON / tools | Yes / Yes / Yes | Yes / Yes / Yes | Yes / Yes / Yes |
| Context or input / max output | 1,050,000 / 128,000 | 1,000,000 / 128,000 | 1,048,576 / 65,536 |
| Task quality | Not verified in shared testing. Compare factual answers, tone and response time using examples in your language. | Not verified in shared testing. Compare factual answers, tone and response time using examples in your language. | Not verified in shared testing. Compare factual answers, tone and response time using examples in your language. |
| Terms | Standard, direct API. Above 272K input tokens, higher rates apply to the entire request. Cache writes: 30 minutes. | Global Claude API, standard mode. Cache writes: 5 minutes. Reasoning can consume output beyond the visible answer. | Paid API, text. Separate input and output limits. Promotional pricing through Dec 31, 2026. Implicit cache reads; explicit storage excluded. |
Scroll the table to see all models.
Open results and download. Optional contact to assess a real application. Pasted text is not submitted; only volumes, requirements and selected models accompany your request.
03 / Explore more optionsView all 11 modelsFilter by provider and profile to build your selection.
Cost for the same scenario.
Lowest estimate first · compare up to 4
Anthropic
Claude Haiku 5.5
- Context
- 1,000,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$4.50/ month
$0.00045 / request
Rates, conditions and sources
claude-haiku-5-5Above 100K input tokens, rates are $0.50 / $2.50 per million. Cache writes: 5 minutes.
Standard USD rates per 1M tokens (batch discount applied to total): input 0.1; output 0.5; cache read 0.01; cache write 0.125. Minimum input for cache: 512 tokens.
OpenAI
GPT-6 Luna
- Context
- 1,050,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$4.50/ month
$0.00045 / request
Rates, conditions and sources
gpt-6-lunaStandard, direct API. Above 272K input tokens, higher rates apply to the entire request. Cache writes: 30 minutes.
Standard USD rates per 1M tokens (batch discount applied to total): input 0.1; output 0.5; cache read 0.01; cache write 0.125. Minimum input for cache: 1,024 tokens.
DeepSeek
DeepSeek V4.1 Flash
- Context
- 1,000,000
- Max output
- 384,000
ImageJSONTools
Monthly estimate
$12.00/ month
$0.0012 / request
Rates, conditions and sources
deepseek-flashPeak rates as a conservative reference. Off-peak rates are half. Aliases and versions can change. Batch not modeled.
Standard USD rates per 1M tokens (batch discount applied to total): input 0.3; output 1.2; cache read 0.006; cache write —.
OpenAI
GPT-4.1 Mini
- Context
- 1,047,576
- Max output
- 32,768
ImageJSONTools
Monthly estimate
$16.00/ month
$0.0016 / request
Rates, conditions and sources
gpt-4.1-miniEarlier generation without a reasoning step. Local o200k_base counts cover text, excluding the API envelope.
Standard USD rates per 1M tokens (batch discount applied to total): input 0.4; output 1.6; cache read 0.1; cache write —.
Gemini 3.8 Flash
- Max input
- 1,048,576
- Max output
- 65,536
ImageJSONTools
Monthly estimate
$33.75/ month
$0.003375 / request
Rates, conditions and sources
gemini-3.8-flashPaid API, text. Separate input and output limits. Promotional pricing through Dec 31, 2026. Implicit cache reads; explicit storage excluded.
Standard USD rates per 1M tokens (batch discount applied to total): input 0.75; output 3.75; cache read 0.075; cache write —. Minimum input for cache: 4,096 tokens. New rate scheduled for 2027-01-01.
DeepSeek
DeepSeek V4 Pro
- Context
- 1,000,000
- Max output
- 384,000
No imageJSONTools
Monthly estimate
$46.20/ month
$0.00462 / request
Rates, conditions and sources
deepseek-v4-proPeak rates as a conservative reference. Off-peak rates are half. Aliases and versions can change. Batch not modeled.
Standard USD rates per 1M tokens (batch discount applied to total): input 1.32; output 3.96; cache read 0.044; cache write —.
Anthropic
Claude Sonnet 5.5
- Context
- 1,000,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$90.00/ month
$0.009 / request
Rates, conditions and sources
claude-sonnet-5-5Global Claude API, standard mode. Cache writes: 5 minutes. Reasoning can consume output beyond the visible answer.
Standard USD rates per 1M tokens (batch discount applied to total): input 2; output 10; cache read 0.1; cache write 2.5. Minimum input for cache: 512 tokens.
Gemini 4 Argon
- Context
- 1,048,576
- Max output
- 65,536
ImageJSONTools
Monthly estimate
$90.00/ month
$0.009 / request
Rates, conditions and sources
gemini-4-argonPromotional pricing through Nov 30, 2026. Implicit caching; default synchronous output limit. Thinking included in output.
Standard USD rates per 1M tokens (batch discount applied to total): input 2; output 10; cache read 0.1; cache write —. Minimum input for cache: 4,096 tokens. New rate scheduled for 2026-12-01.
OpenAI
GPT-6.1 Sol
- Context
- 1,050,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$90.00/ month
$0.009 / request
Rates, conditions and sources
gpt-6.1-solStandard, direct API. Above 272K input tokens, higher rates apply to the entire request. Cache writes: 30 minutes.
Standard USD rates per 1M tokens (batch discount applied to total): input 2; output 10; cache read 0.1; cache write 2.5. Minimum input for cache: 1,024 tokens.
Anthropic
Claude Opus 5.5
- Context
- 1,000,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$180.00/ month
$0.02 / request
Rates, conditions and sources
claude-opus-5-5Global Claude API, standard mode. Cache writes: 5 minutes. Reasoning can consume output beyond the visible answer.
Standard USD rates per 1M tokens (batch discount applied to total): input 4; output 20; cache read 0.2; cache write 5. Minimum input for cache: 512 tokens.
OpenAI
GPT-6 Astra
- Context
- 1,050,000
- Max output
- 128,000
ImageJSONTools
Monthly estimate
$450.00/ month
$0.05 / request
Rates, conditions and sources
gpt-6-astraStandard, direct API. Above 272K input tokens, higher rates apply to the entire request. Cache writes: 30 minutes.
Standard USD rates per 1M tokens (batch discount applied to total): input 10; output 50; cache read 1; cache write 12.5. Minimum input for cache: 1,024 tokens.
3/4 selectedView comparison
What is included?
Input and output tokens, modeled cache reads and writes, monthly volume and batch without caching. Text estimate via direct API, excluding chat subscriptions, media, tools, infrastructure, taxes and negotiated prices.
Are these models at the same level?
Grouping is editorial, based on provider positioning. The same profile and context window do not guarantee equal quality. Comparable tests require the same task, rubric, versions, tools and reasoning budget.
Is counting exact?
o200k_base counts text in that encoding. Billing depends on each API’s tokenizer and message format. For a real workload, use provider-reported usage counts in manual mode.
How are prices updated?
Manually reviewed catalog with official links and a checked date. Announced future changes apply on their scheduled date. Estimates use the page calculation date; confirm current provider terms.
To turn estimates into an application, explore our AI consulting services.
Planning infrastructure too? Compare VPS and cloud prices.