GPT-6.1 Sol
Text token rates. Cache writes $2.50 / $5. Fast 2×; Batch/Flex 50% lower. Regional processing +10% where available. Tool calls are separate.
Price effective date: not stated in the verified evidence.
Model API pricing
Compare official token prices, billing conditions and verified history. ChatGPT and Codex subscriptions are billed separately from API usage.
Last verified · · Each row retains its own successful check.
Anthropic
| Model / tier | Input | Output | Cached input | Official source / last verified |
|---|---|---|---|---|
Claude Haiku 4.5claude-haiku-4-5-20251001Standard · 200K contextActive | $1 | $5 | $0.1 | Official pricing |
Claude Opus 5.5claude-opus-5-5Standard · up to 1M contextActive | $4 | $20 | $0.2 | Official pricing |
Claude Sonnet 5.5claude-sonnet-5-5Standard · up to 1M contextActive | $2 | $10 | $0.2 | Official pricing |
| Model / tier | Input | Output | Cached input | Official source / last verified |
|---|---|---|---|---|
Gemini 3.1 Pro Previewgemini-3.1-pro-previewStandard · input ≤200KActive · Preview | $2 | $12 | $0.2 | Official pricing |
Gemini 3.1 Pro Previewgemini-3.1-pro-previewStandard · input >200KActive · Preview | $4 | $18 | $0.4 | Official pricing |
Gemini 3.5 Flash-Litegemini-3.5-flash-liteStandard · text / image / video / audio inputActive | $0.3 | $2.5 | $0.03 | Official pricing |
Gemini 3.8 Flashgemini-3.8-flashStandard · text / image / video inputActive | $0.75 | $3.75 | $0.075 | Official pricing |
OpenAI
| Model / tier | Input | Output | Cached input | Official source / last verified |
|---|---|---|---|---|
GPT-6 Astragpt-6-astraStandard · input ≤272KActive | $10 | $50 | $1 | Official pricing |
GPT-6 Astragpt-6-astraStandard · input >272KActive | $20 | $75 | $2 | Official pricing |
GPT-6.1 Solgpt-6.1-solStandard · input ≤272KActive | $2 | $10 | $0.1 | Official pricing |
GPT-6.1 Solgpt-6.1-solStandard · input >272KActive | $4 | $15 | $0.2 | Official pricing |
Billing conditions
GPT-6 Astra
Cache writes $12.50 / $25. Fast 2×; Batch/Flex 50% lower. Ultrafast short-context input/output $60/$300: see the linked API pricing policy, not a subscription allowance.
Price effective date: not stated in the verified evidence.
Claude Opus 5.5
Cache write: $5 (5m), $8 (1h). Batch input/output 50% lower. US-only processing 1.1×. Fast preview $8/$40; no Batch.
Price effective date: not stated in the verified evidence.
Claude Sonnet 5.5
Cache write: $2.50 (5m), $4 (1h). Batch input/output 50% lower. US-only processing 1.1×.
Price effective date: not stated in the verified evidence.
Claude Haiku 4.5
Cache write: $1.25 (5m), $2 (1h). Batch input/output 50% lower. Cache storage duration and write charges are separate from reads.
Price effective date: not stated in the verified evidence.
Gemini 3.8 Flash
Output includes thinking tokens. Cache storage $0.50 / 1M token-hour. Batch $0.375/$1.875, cache $0.0375. Current introductory rates end 2026-12-31; announced 2027 rates are not current prices.
Price effective date: not stated in the verified evidence.
Gemini 3.5 Flash-Lite
Output includes thinking tokens. Cache storage $1 / 1M token-hour. Batch $0.15/$1.25, cached input $0.02 (officially listed; not calculated as half). The listed Standard and Batch input rates also apply to audio.
Price effective date: not stated in the verified evidence.
Gemini 3.1 Pro Preview
Output includes thinking tokens. Cache storage $4.50 / 1M token-hour. Batch input/output 50% lower; cached input stays $0.20/$0.40. Preview model.
Price effective date: not stated in the verified evidence.
Cached reads, cache writes and storage are different charges. Tool calls, audio and third-party hosting can add separate costs; these text-token prices are not a complete usage bill.
Pricing timeline
Initial records are observations, not price cuts. Effective dates remain unknown unless an official source states them. Verification without a price change does not add a history event.
- Gemini 3.5 Flash-Lite · Billing terms changeStandard · text / image / video / audio input: $0.3 / $2.5 · Cached input: $0.03
Output includes thinking tokens. Cache storage $1 / 1M token-hour. Batch $0.15/$1.25, cached input $0.02 (officially listed; not calculated as half). The listed Standard and Batch input rates also apply to audio.
Official source - GPT-6 Astra · First verified recordStandard · input ≤272K: $10 / $50 · Cached input: $1Standard · input >272K: $20 / $75 · Cached input: $2Official source
- GPT-6.1 Sol · First verified recordStandard · input ≤272K: $2 / $10 · Cached input: $0.1Standard · input >272K: $4 / $15 · Cached input: $0.2Official source
- Claude Haiku 4.5 · First verified recordStandard · 200K context: $1 / $5 · Cached input: $0.1Official source
- Claude Opus 5.5 · First verified recordStandard · up to 1M context: $4 / $20 · Cached input: $0.2Official source
- Claude Sonnet 5.5 · First verified recordStandard · up to 1M context: $2 / $10 · Cached input: $0.2Official source
- Gemini 3.1 Pro Preview · First verified recordStandard · input ≤200K: $2 / $12 · Cached input: $0.2Standard · input >200K: $4 / $18 · Cached input: $0.4Official source
- Gemini 3.5 Flash-Lite · First verified recordStandard · text / image / video input: $0.3 / $2.5 · Cached input: $0.03Official source
- Gemini 3.8 Flash · First verified recordStandard · text / image / video input: $0.75 / $3.75 · Cached input: $0.075Official source
Announced future prices
gemini-3.8-flash · : $1.5 / $7.5 · Not the current rate · Official announcement