Documents
A 50-page document translated in full — the rare job where output tokens outweigh input.
Pricing verified:
Between $0.0068 and $0.425 per run, depending on the model. That is a 62× spread for the same job.
50 pages of dense prose at ~500 words per page ≈ 32,500 input tokens. Translated output runs roughly 10% longer than the source in most European language pairs ≈ 36,000 output tokens.
Cost = (input tokens ÷ 1M × input price) + (output tokens ÷ 1M × output price). Prices come from our daily-verified model data. Batch and cached-input discounts are not applied — they only ever make these numbers smaller.
| Model | Provider | Per run | At 40 documents a month | Context |
|---|---|---|---|---|
| Gemma 2 9Bneeds chunking | $0.0042 | $0.169 | 8k tokens | |
| Llama 3.1 8B Instructneeds chunking | Meta | $0.0045 | $0.180 | 16k tokens |
| Ministral 3 3B 2512 | Mistral | $0.0068 | $0.274 | 131k tokens |
| Llama 3.2 1B Instructneeds chunking | Meta | $0.0081 | $0.323 | 60k tokens |
| gpt-oss-safeguard-20b | OpenAI | $0.0095 | $0.379 | 131k tokens |
| Llama 3 8B Instructneeds chunking | Meta | $0.0096 | $0.384 | 8k tokens |
| Mistral Small 3.2 24B | Mistral | $0.0096 | $0.385 | 128k tokens |
| Ministral 3 8B 2512 | Mistral | $0.010 | $0.411 | 262k tokens |
| Mistral Nemo | Mistral | $0.010 | $0.411 | 131k tokens |
| Mistral 7B Instruct v0.1needs chunking | Mistral | $0.010 | $0.417 | 3k tokens |
| Llama Guard 4 12B | Meta | $0.012 | $0.493 | 164k tokens |
| Gemini 2.0 Flash Lite | $0.013 | $0.529 | 1.0M tokens | |
| Ministral 3 14B 2512 | Mistral | $0.014 | $0.548 | 262k tokens |
| Devstral Small 1.1 | Mistral | $0.014 | $0.562 | 131k tokens |
| Mistral Small Creativeneeds chunking | Mistral | $0.014 | $0.562 | 33k tokens |
| Voxtral Small 24B 2507needs chunking | Mistral | $0.014 | $0.562 | 32k tokens |
| Mistral Small 3.1 | Mistral | $0.014 | $0.562 | 128k tokens |
| DeepSeek V4-Flash | DeepSeek | $0.015 | $0.585 | 1M tokens |
| GPT-5 Nano | OpenAI | $0.016 | $0.641 | 400k tokens |
| Llama Guard 3 8B | Meta | $0.017 | $0.667 | 131k tokens |
| GPT-4.1 Nano | OpenAI | $0.018 | $0.706 | 1.0M tokens |
| Gemini 2.0 Flash | $0.018 | $0.706 | 1.0M tokens | |
| Gemini 2.5 Flash Lite | $0.018 | $0.706 | 1.0M tokens | |
| Gemini 2.5 Flash Lite Preview 09-2025 | $0.018 | $0.706 | 1.0M tokens | |
| Gemma 4 26B A4B | $0.019 | $0.745 | 262k tokens | |
| Qwen 3.8 Flash | Alibaba | $0.022 | $0.885 | 991k tokens |
| GLM-5.3 Flash | Z.ai | $0.023 | $0.915 | 1M tokens |
| Llama 3.2 11B Vision Instruct | Meta | $0.024 | $0.945 | 131k tokens |
| GPT-4o Mini | OpenAI | $0.026 | $1.06 | 128k tokens |
| Mistral Small 4 | Mistral | $0.026 | $1.06 | 262k tokens |
| Llama 3.1 70B Instruct | Meta | $0.027 | $1.10 | 131k tokens |
| Grok 3 Mini | xAI | $0.028 | $1.11 | 131k tokens |
| Grok 3 Mini Beta | xAI | $0.028 | $1.11 | 131k tokens |
| Sabaneeds chunking | Mistral | $0.028 | $1.12 | 33k tokens |
| Mistral Small 3needs chunking | Mistral | $0.031 | $1.26 | 33k tokens |
| Mixtral 8x7B Instructneeds chunking | Mistral | $0.037 | $1.48 | 33k tokens |
| Codestral 2508 | Mistral | $0.042 | $1.69 | 256k tokens |
| Llama 3 70B Instructneeds chunking | Meta | $0.043 | $1.73 | 8k tokens |
| Gemma 2 27Bneeds chunking | $0.045 | $1.78 | 8k tokens | |
| DeepSeek V4-Pro | DeepSeek | $0.045 | $1.82 | 1M tokens |
| Gemma 4 31B | $0.048 | $1.90 | 262k tokens | |
| DeepSeek V3 | DeepSeek | $0.048 | $1.94 | 128k tokens |
| GPT-5.6 Luna | OpenAI | $0.050 | $1.99 | 1.1M tokens |
| Claude 3 Haiku | Anthropic | $0.053 | $2.13 | 200k tokens |
| Llama 4 Scout | Meta | $0.059 | $2.38 | 512k tokens |
| Grok Code Fast 1 | xAI | $0.060 | $2.42 | 256k tokens |
| Gemini 3 Flash Preview | $0.062 | $2.48 | 1.0M tokens | |
| Muse Glimmer 30B | Meta | $0.065 | $2.61 | 131k tokens |
| GPT-3.5 Turboneeds chunking | OpenAI | $0.070 | $2.81 | 16k tokens |
| Mistral Large 3 2512 | Mistral | $0.070 | $2.81 | 262k tokens |
| GPT-4.1 Mini | OpenAI | $0.071 | $2.82 | 1.0M tokens |
| Llama 4 Maverick | Meta | $0.077 | $3.08 | 256k tokens |
| GPT-5 Mini | OpenAI | $0.080 | $3.21 | 400k tokens |
| GPT-5.1-Codex-Mini | OpenAI | $0.080 | $3.21 | 400k tokens |
| Devstral 2 2512 | Mistral | $0.085 | $3.40 | 262k tokens |
| Devstral Medium | Mistral | $0.085 | $3.40 | 131k tokens |
| Mistral Medium 3 | Mistral | $0.085 | $3.40 | 131k tokens |
| Mistral Medium 3.1 | Mistral | $0.085 | $3.40 | 131k tokens |
| DeepSeek R1 | DeepSeek | $0.097 | $3.87 | 128k tokens |
| Gemini 2.5 Flash | $0.100 | $3.99 | 1.0M tokens | |
| Gemini 3.5 Flash-Lite | $0.100 | $3.99 | 1.0M tokens | |
| Nano Banana (Gemini 2.5 Flash Image)needs chunking | $0.100 | $3.99 | 33k tokens | |
| GPT-3.5 Turbo (older v0613)needs chunking | OpenAI | $0.104 | $4.18 | 4k tokens |
| GPT Audio Mini | OpenAI | $0.106 | $4.24 | 128k tokens |
| GPT-3.5 Turbo Instructneeds chunking | OpenAI | $0.121 | $4.83 | 4k tokens |
| Gemini 3.1 Flash | $0.124 | $4.97 | 1M tokens | |
| Codestral 25.01 | Mistral | $0.126 | $5.06 | 256k tokens |
| GPT-5 Image Mini | OpenAI | $0.153 | $6.13 | 400k tokens |
| Gemini 3.6 Flash | $0.159 | $6.38 | 1.0M tokens | |
| Gemini 3.7 Flash | $0.159 | $6.38 | 1.0M tokens | |
| Claude 3.5 Haiku | Anthropic | $0.170 | $6.80 | 200k tokens |
| Claude 4 Haiku | Anthropic | $0.170 | $6.80 | 200k tokens |
| Kimi K2.7 Code | Moonshot | $0.175 | $7.00 | 256k tokens |
| Muse Spark | Meta | $0.194 | $7.74 | 1.0M tokens |
| o3 Mini | OpenAI | $0.194 | $7.77 | 200k tokens |
| o3 Mini High | OpenAI | $0.194 | $7.77 | 200k tokens |
| o4 Mini | OpenAI | $0.194 | $7.77 | 200k tokens |
| o4 Mini High | OpenAI | $0.194 | $7.77 | 200k tokens |
| GLM-5.2 | Z.ai | $0.204 | $8.16 | 1M tokens |
| GLM-5.3 | Z.ai | $0.204 | $8.16 | 1M tokens |
| GPT-5.2 Mini | OpenAI | $0.212 | $8.47 | 128k tokens |
| Claude Haiku 4.5 | Anthropic | $0.212 | $8.50 | 200k tokens |
| GPT-3.5 Turbo 16kneeds chunking | OpenAI | $0.241 | $9.66 | 16k tokens |
| Grok 4 | xAI | $0.281 | $11.24 | 2M tokens |
| Grok 4.5 | xAI | $0.281 | $11.24 | 500k tokens |
| Grok 4.6 | xAI | $0.281 | $11.24 | 500k tokens |
| Mixtral 8x22B Instructneeds chunking | Mistral | $0.281 | $11.24 | 66k tokens |
| Pixtral Large 2411 | Mistral | $0.281 | $11.24 | 131k tokens |
| Qwen 3.8 Max | Alibaba | $0.281 | $11.24 | 1M tokens |
| Mistral Medium 3.5 | Mistral | $0.319 | $12.75 | 256k tokens |
| Qwen 3.7 Max | Alibaba | $0.351 | $14.05 | 1M tokens |
| GPT-4.1 | OpenAI | $0.353 | $14.12 | 1.0M tokens |
| o3 | OpenAI | $0.353 | $14.12 | 200k tokens |
| o4 Mini Deep Research | OpenAI | $0.353 | $14.12 | 200k tokens |
| Gemini 3.5 Flash | $0.373 | $14.91 | 1.0M tokens | |
| GPT-5 | OpenAI | $0.401 | $16.02 | 400k tokens |
| GPT-5 Chat | OpenAI | $0.401 | $16.02 | 128k tokens |
| GPT-5 Codex | OpenAI | $0.401 | $16.02 | 400k tokens |
| GPT-5.1 | OpenAI | $0.401 | $16.02 | 400k tokens |
| GPT-5.1 Chat | OpenAI | $0.401 | $16.02 | 128k tokens |
| GPT-5.1-Codex | OpenAI | $0.401 | $16.02 | 400k tokens |
| GPT-5.1-Codex-Max | OpenAI | $0.401 | $16.02 | 400k tokens |
| Gemini 2.5 Pro | $0.401 | $16.02 | 1.0M tokens | |
| Gemini 2.5 Pro Preview 05-06 | $0.401 | $16.02 | 1.0M tokens | |
| Gemini 2.5 Pro Preview 06-05 | $0.401 | $16.02 | 1.0M tokens | |
| Mistral Large 2 | Mistral | $0.421 | $16.86 | 128k tokens |
| Claude Sonnet 5 | Anthropic | $0.425 | $17.00 | 1M tokens |
| GPT-5.6 Sol | OpenAI | $0.425 | $17.00 | 1.1M tokens |
| GPT Audio | OpenAI | $0.441 | $17.65 | 128k tokens |
| GPT-4o | OpenAI | $0.441 | $17.65 | 128k tokens |
| GPT-5.6 Terra | OpenAI | $0.497 | $19.88 | 1.1M tokens |
| Gemini 3.1 Pro | $0.497 | $19.88 | 2M tokens | |
| Nano Banana Pro (Gemini 3 Pro Image Preview)needs chunking | $0.497 | $19.88 | 66k tokens | |
| GPT-5.2 | OpenAI | $0.561 | $22.44 | 200k tokens |
| GPT-5.3 Chat | OpenAI | $0.561 | $22.44 | 128k tokens |
| GPT-5.3-Codex | OpenAI | $0.561 | $22.44 | 400k tokens |
| GPT-5.4 | OpenAI | $0.621 | $24.85 | 272k tokens |
| Claude 3.7 Sonnet (thinking) | Anthropic | $0.637 | $25.50 | 200k tokens |
| Claude Sonnet 4 | Anthropic | $0.637 | $25.50 | 200k tokens |
| Claude Sonnet 4.5 | Anthropic | $0.637 | $25.50 | 1M tokens |
| Claude Sonnet 4.6 | Anthropic | $0.637 | $25.50 | 1M tokens |
| Grok 3 | xAI | $0.637 | $25.50 | 131k tokens |
| Grok 3 Beta | xAI | $0.637 | $25.50 | 131k tokens |
| Kimi K3 | Moonshot | $0.637 | $25.50 | 1M tokens |
| GPT-5 Image | OpenAI | $0.685 | $27.40 | 400k tokens |
| Claude Opus 4.5 | Anthropic | $1.06 | $42.50 | 200k tokens |
| Claude Opus 4.6 | Anthropic | $1.06 | $42.50 | 1M tokens |
| Claude Opus 4.7 | Anthropic | $1.06 | $42.50 | 1M tokens |
| Claude Opus 4.8 | Anthropic | $1.06 | $42.50 | 1M tokens |
| Claude Opus 5 | Anthropic | $1.06 | $42.50 | 1M tokens |
| GPT-5.5 | OpenAI | $1.24 | $49.70 | 1M tokens |
| Claude 3.5 Sonnet | Anthropic | $1.27 | $51.00 | 200k tokens |
| GPT-4 Turbo | OpenAI | $1.40 | $56.20 | 128k tokens |
| GPT-4 Turbo (older v1106) | OpenAI | $1.40 | $56.20 | 128k tokens |
| GPT-4 Turbo Preview | OpenAI | $1.40 | $56.20 | 128k tokens |
| o3 Deep Research | OpenAI | $1.76 | $70.60 | 200k tokens |
| Claude Fable 5 | Anthropic | $2.13 | $85.00 | 1M tokens |
| Claude Fable 5.1 | Anthropic | $2.13 | $85.00 | 1M tokens |
| Claude Mythos 5 | Anthropic | $2.13 | $85.00 | 1M tokens |
| Claude Mythos 5.1 | Anthropic | $2.13 | $85.00 | 1M tokens |
| GPT-6 Astra | OpenAI | $2.13 | $85.00 | 1.1M tokens |
| o1 | OpenAI | $2.65 | $106 | 200k tokens |
| GPT-4needs chunking | OpenAI | $3.13 | $125 | 8k tokens |
| GPT-4 (older v0314)needs chunking | OpenAI | $3.13 | $125 | 8k tokens |
| Claude Opus 4 | Anthropic | $3.19 | $128 | 200k tokens |
| Claude Opus 4.1 | Anthropic | $3.19 | $128 | 200k tokens |
| o3 Pro | OpenAI | $3.53 | $141 | 200k tokens |
| GPT-5 Pro | OpenAI | $4.81 | $192 | 400k tokens |
| o1-pro | OpenAI | $26.47 | $1,059 | 200k tokens |
“Needs chunking” means the model's context window can't hold this job in a single pass, so the real cost is higher than the figure shown and quality usually suffers.
Translation is the one common task where you pay full output rates on the entire document, and output tokens are typically 3–5× the price of input. That inverts the usual ranking: models that look cheap on input-heavy work can be the expensive option here.
$0.0068 on Ministral 3 3B 2512, the cheapest capable option, rising to $0.425 on Claude Sonnet 5 at the top end. DeepSeek V4-Pro is the value pick at $0.045 per run. The job is priced at 32,500 input and 36,000 output tokens — see the working below.
50 pages of dense prose at ~500 words per page ≈ 32,500 input tokens. Translated output runs roughly 10% longer than the source in most European language pairs ≈ 36,000 output tokens.
$0.274 a month on Ministral 3 3B 2512, $1.82 on DeepSeek V4-Pro, and $17.00 on Claude Sonnet 5. Batch APIs typically halve these figures for work that can wait, and prompt caching cuts the input side further when the same context is reused.
Translation is the one common task where you pay full output rates on the entire document, and output tokens are typically 3–5× the price of input. That inverts the usual ranking: models that look cheap on input-heavy work can be the expensive option here.
Yes. Every figure on this page is computed from our model pricing data, which is checked daily against each provider's official pricing page. When a provider changes a price, these numbers change with it.
Newsletter
Pricing changes, new model releases, and updated recommendations — delivered when it matters.
No spam. Useful updates only. Affiliate disclosures always clearly labeled.