Together AI API pricing, checked at the source
Serverless prices for open models on Together AI, read off Together's own pricing page with a link on every row. Estimate a monthly bill below.
Together AI API prices per 1M tokens
Standard prices in USD. Each row links to Together AI's pricing page and shows the day we checked it.
| Model | Input | Cached input | Output | Notes | Source |
|---|---|---|---|---|---|
| Kimi K3 | $3.00 | $0.30 | $15.00 | Sourcechecked 2026-10-25 | |
| Qwen3.8-2.4T-A95B | $2.00 | $0.25 | $6.00 | The docs models table lists cached input at $0.50 for this model; the pricing page shows $0.25. | Sourcechecked 2026-10-25 |
| GLM-5.3 | $1.40 | $0.26 | $4.40 | Sourcechecked 2026-10-25 | |
| DeepSeek V4 Pro 0813 | $1.32 | $0.13 | $3.96 | Sourcechecked 2026-10-25 | |
| Muse Spark Open | $1.25 | $0.15 | $4.25 | Sourcechecked 2026-10-25 | |
| MiniMax M3 | $0.30 | $0.06 | $1.20 | Sourcechecked 2026-10-25 | |
| DeepSeek V4.1 Flash 1017 | $0.30 | $0.006 | $1.20 | Sourcechecked 2026-10-25 | |
| GLM-5.3-Flash | $0.15 | $0.03 | $0.50 | Sourcechecked 2026-10-25 | |
| gpt-oss-120B | $0.15 | none listed | $0.60 | Sourcechecked 2026-10-25 | |
| Qwen3.5 9B | $0.17 | none listed | $0.25 | Sourcechecked 2026-10-25 |
Terms on this page: Token · Prompt caching · Context window · Batch API
Estimate your Together AI bill
Enter your monthly tokens and how they split between new input, cached input and output. The result updates as you type.
Write 40k, 250M or 5B. Input, cached input and output tokens together.
100M tokens a month costs from $9.79 (GLM-5.3-Flash) to $203.40 (Kimi K3).
| Model | Per month | Per 1M tokens at this split |
|---|---|---|
| GLM-5.3-Flash | $9.79 | $0.0979 |
| gpt-oss-120B | $17.25 | $0.1725 |
| Qwen3.5 9B | $17.40 | $0.174 |
| DeepSeek V4.1 Flash 1017 | $17.45 | $0.17448 |
| MiniMax M3 | $20.58 | $0.2058 |
| DeepSeek V4 Pro 0813 | $76.18 | $0.7618 |
| Muse Spark Open | $76.20 | $0.762 |
| GLM-5.3 | $88.88 | $0.8888 |
| Qwen3.8-2.4T-A95B | $118.50 | $1.185 |
| Kimi K3 | $203.40 | $2.034 |
Standard prices in USD on 2026-10-25, before tax. Long-prompt tiers are left out; see the price table.
Together AI speech to text prices
Prices as the vendor lists them, with a per-hour column so rows compare. Add-ons such as speaker labels cost extra; see the billing notes.
| Model | Price as listed | Per hour | Notes | Source |
|---|---|---|---|---|
| NVIDIA Parakeet TDT 0.6B v3 | $0.0015 per minute | $0.09 | Sourcechecked 2026-10-25 | |
| Whisper Large v3 | $0.0015 per minute | $0.09 | Sourcechecked 2026-10-25 |
Write 40, 250 or 5k.
100 hours of audio a month costs from $9.00 (NVIDIA Parakeet TDT 0.6B v3) to $9.00 (Whisper Large v3).
| Model | Per month | Per hour |
|---|---|---|
| NVIDIA Parakeet TDT 0.6B v3 | $9.00 | $0.09 |
| Whisper Large v3 | $9.00 | $0.09 |
List prices in USD on 2026-10-25, before tax, free tiers and add-ons. Rows billed another way are in the price table only.
Together AI text to speech prices
Prices as the vendor lists them, with a per-1M-characters column so rows compare.
| Model | Price as listed | Per 1M characters | Notes | Source |
|---|---|---|---|---|
| Kokoro-82M TTS | $4.00 per 1M characters | $4.00 | Sourcechecked 2026-10-25 | |
| Cartesia Sonic-3 | $65.00 per 1M characters | $65.00 | Sourcechecked 2026-10-25 |
Write 500k, 10M or 1B.
10M characters a month costs from $40.00 (Kokoro-82M TTS) to $650.00 (Cartesia Sonic-3).
| Model | Per month | Per 1M characters |
|---|---|---|
| Kokoro-82M TTS | $40.00 | $4.00 |
| Cartesia Sonic-3 | $650.00 | $65.00 |
List prices in USD on 2026-10-25, before tax, free tiers and add-ons. Rows billed another way are in the price table only.
Together AI image generation prices
Prices per image where the vendor lists one. Rows billed per megapixel or per token show how they bill instead.
| Model | Price as listed | Per image | Notes | Source |
|---|---|---|---|---|
| FLUX.2 [dev] | $0.0154 per image | $0.0154 | Estimate from the lowest settings; real cost varies. | Sourcechecked 2026-10-25 |
| ByteDance Seedream 4.0 | $0.03 per image | $0.03 | Read from the docs image table, which lists the unit as image. | Sourcechecked 2026-10-25 |
| FLUX.2 [pro] | $0.03 per image | $0.03 | Estimate from the lowest settings; real cost varies with resolution and options. | Sourcechecked 2026-10-25 |
| GPT Image 2 | $0.053 per image | $0.053 | Estimate from the lowest settings; real cost varies. | Sourcechecked 2026-10-25 |
| Nano Banana Pro (Gemini 3 Pro Image) | $0.134 per image | $0.134 | At 1K or 2K; docs say about $0.24 at 4K. | Sourcechecked 2026-10-25 |
| FLUX1.1 [pro] | $0.04 per megapixel | billed per megapixel | Sourcechecked 2026-10-25 | |
| FLUX.2 [max] | $0.07 per megapixel | billed per megapixel | At the default 50 steps; more steps cost proportionally more. | Sourcechecked 2026-10-25 |
Write 500, 10k or 1M.
1k images a month costs from $15.40 (FLUX.2 [dev]) to $134.00 (Nano Banana Pro (Gemini 3 Pro Image)).
| Model | Per month | Per image |
|---|---|---|
| FLUX.2 [dev] | $15.40 | $0.0154 |
| ByteDance Seedream 4.0 | $30.00 | $0.03 |
| FLUX.2 [pro] | $30.00 | $0.03 |
| GPT Image 2 | $53.00 | $0.053 |
| Nano Banana Pro (Gemini 3 Pro Image) | $134.00 | $0.134 |
List prices in USD on 2026-10-25, before tax, free tiers and add-ons. Rows billed another way are in the price table only.
How the bill is worked out
Batch discount is narrow today
Cached input is cheaper
Region premium
Dedicated is billed by GPU time
Together AI API pricing questions
How much does the Together AI API cost?
Together AI bills its language models per million tokens, with separate prices for input, cached input and output. On 2026-10-25, standard prices ran from $0.15 (GLM-5.3-Flash) to $3.00 (Kimi K3) per 1M input tokens, and $0.25 (Qwen3.5 9B) to $15.00 (Kimi K3) per 1M output tokens. Speech to text runs from $0.09 per hour. Text to speech runs from $4.00 (Kokoro-82M TTS) to $65.00 (Cartesia Sonic-3) per 1M characters. Image generation runs from $0.0154 (FLUX.2 [dev]) to $0.134 (Nano Banana Pro (Gemini 3 Pro Image)) per image.
Does Together AI offer a batch discount?
Its pricing page lists no batch price for the rows here. The Batch API says up to 50% off, but the docs list only openai/whisper-large-v3 as discounted. Models not listed run at standard rates.
How does prompt caching change a Together AI bill?
Cached tokens bill at the model's cached input rate. Writing to the cache is free. Models with no cached price bill all input at the normal rate.
Where do these prices come from?
From Together AI's own pricing page, https://www.together.ai/pricing, checked on 2026-10-25. Every row links to it. When a price cannot be confirmed, the row says to check the vendor's page instead of showing a number.
Does Together AI have a batch discount?
Its Batch API says up to 50% off, but its docs list only whisper-large-v3 as discounted today. Other models run at the standard rates in the table.
Compare and plan
Building with the Together AI API? OceanLab builds AI into business workflows, with people kept in charge. AI automation →
