AI API price index · open-model host · read 16 Oct 2026
Fireworks AI API pricing
29 prices as Fireworks AI publishes them — 21 text models, 3 embeddings, 5 hardware time — read on its own pricing page on 16 Oct 2026. Its current text models cost $0.20 to $15.00 per million output tokens.
Free tier The pricing page states no free tier or credit amount ('Pay per token with pre-paid, usage-based billing. Simply add a payment method to buy credits'). The official billing FAQ is titled 'What happens when I finish my $1 credit?' and says that without a payment method the account is suspended once that $1 credit is used — i.e. a $1 starting credit is implied, not stated as an offer.
What to know before you build on Fireworks AI
- Serverless text prices are on docs.fireworks.ai/serverless/pricing; the main pricing page links there.
- Only 16 models are priced by name. Any other text model is priced by size band, the same rate in and out, from $0.10 to $1.20 per million tokens.
- US-served variants cost 50% more than the default, and Priority mode 25% more.
- Batch inference is billed at 50% of serverless prices.
- No free tier is offered on the pricing page; the billing FAQ refers to a $1 starting credit.
Against the other ten
Across the index, 85 current text models run from $0.10 to $50.00 per million output tokens — 500×. Fireworks AI’s cheapest, NVIDIA Nemotron 3.5 Lightning 30B A3B at $0.20, has 5 cheaper text models elsewhere; its dearest, Kimi K3, costs $15.00. A cheaper model is a different model, not the same one for less — compare capability before price.
The same model elsewhere
Open-weight models Fireworks AI sells that other hosts price too — the only like-for-like comparison in the index.
- DeepSeek V4.1 Flash — Fireworks AI $0.30 / $1.20 · Together AI $0.30 / $1.20 — the same everywhere
- GLM 5.3 — Fireworks AI $1.40 / $4.40 · Mistral AI $1.40 / $4.40 · Together AI $1.40 / $4.40 — the same everywhere
- GLM 5.3 Flash — Fireworks AI $0.15 / $0.50 · Together AI $0.15 / $0.50 — the same everywhere
- gpt-oss-120b — Fireworks AI $0.15 / $0.60 · Groq $0.15 / $0.60 · Together AI $0.15 / $0.60 — the same everywhere
- Kimi K3 — Fireworks AI $3.00 / $15.00 · Together AI $3.00 / $15.00 — the same everywhere
- MiniMax M3 — Fireworks AI $0.30 / $1.20 · Together AI $0.30 / $1.20 — the same everywhere
Input / output per 1M tokens, standard tier.
Text models
| Model | Input | Cached input | Output | Note |
|---|---|---|---|---|
| Ember-1serverless | $3.00 | $0.30 | $15.00 | Standard mode; Priority $3.75 / $0.375 / $18.75 |
| Kimi K3serverless | $3.00 | $0.30 | $15.00 | Standard mode; Priority $3.75 / $0.375 / $18.75 |
| Kimi K3 Fastserverless | $4.50 | $0.45 | $22.50 | Fast serving path; Priority not available · Not ranked: faster serving path of a model ranked under its default name. |
| Kimi K3 (US)serverless | $4.50 | $0.45 | $22.50 | US variant, Standard mode; Priority $5.625 / $0.5625 / $28.125 · Not ranked: US-served variant of a model ranked under its default name. |
| DeepSeek V4.1 Flashserverless | $0.30 | $0.006 | $1.20 | Standard mode; Priority $0.375 / $0.0075 / $1.50 |
| DeepSeek V4.1 Flash (US)serverless | $0.45 | $0.009 | $1.80 | US variant, Standard mode; Priority $0.5625 / $0.01125 / $2.25 · Not ranked: US-served variant of a model ranked under its default name. |
| GLM 5.3 Flashserverless | $0.15 | $0.03 | $0.50 | Standard mode; Priority $0.1875 / $0.0375 / $0.625 |
| GLM 5.3 Flash (US)serverless | $0.225 | $0.045 | $0.75 | US variant, Standard mode; Priority $0.28125 / $0.05625 / $0.9375 · Not ranked: US-served variant of a model ranked under its default name. |
| GLM 5.3serverless | $1.40 | $0.26 | $4.40 | Standard mode; Priority $1.75 / $0.325 / $5.50 |
| GLM 5.3 (US)serverless | $2.10 | $0.39 | $6.60 | US variant, Standard mode; Priority $2.625 / $0.4875 / $8.25 · Not ranked: US-served variant of a model ranked under its default name. |
| GLM 5.3 Fastserverless | $2.10 | $0.39 | $6.60 | Fast serving path; Priority not available · Not ranked: faster serving path of a model ranked under its default name. |
| Qwen 3.8 Maxserverless | $2.00 | $0.25 | $6.00 | Standard mode; Priority $3.00 / $0.375 / $9.00 |
| MiniMax M3serverless | $0.30 | $0.06 | $1.20 | Standard mode; Priority $0.45 / $0.09 / $1.80 |
| OpenAI GPT OSS 120Bserverless | $0.15 | $0.015 | $0.60 | Standard mode; Priority $0.18 / $0.018 / $0.72 |
| NVIDIA Nemotron 3.5 Lightning 30B A3Bserverless | $0.05 | $0.01 | $0.20 | Standard mode; Priority $0.0625 / $0.0125 / $0.25 |
| NVIDIA Nemotron 3 Ultra (Preview)serverless | $0.60 | $0.12 | $2.40 | Standard mode; Priority $0.75 / $0.15 / $3.00 |
| Less than 4B parametersserverless | $0.10 | — | $0.10 | size band; one price applies uniformly to input and output, no cached rate · Not ranked: a size band, not a named model. |
| 4B – 16B parametersserverless | $0.20 | — | $0.20 | size band; one price applies uniformly to input and output, no cached rate · Not ranked: a size band, not a named model. |
| More than 16B parametersserverless | $0.90 | — | $0.90 | size band; one price applies uniformly to input and output, no cached rate · Not ranked: a size band, not a named model. |
| MoE up to 56B parameters (e.g. Mixtral 8x7B)serverless | $0.50 | — | $0.50 | size band; one price applies uniformly to input and output, no cached rate · Not ranked: a size band, not a named model. |
| MoE 56.1B – 176B parameters (e.g. DBRX, Mixtral 8x22B)serverless | $1.20 | — | $1.20 | size band; one price applies uniformly to input and output, no cached rate · Not ranked: a size band, not a named model. |
Embeddings
| Model | Input | Cached input | Note |
|---|---|---|---|
| up to 150Mserverless | $0.008 | — | size band (base model parameter count); billed on input tokens only · Not ranked: a size band, not a named model. |
| 150M – 350Mserverless | $0.016 | — | size band (base model parameter count); billed on input tokens only · Not ranked: a size band, not a named model. |
| Qwen3 8Bserverless | $0.10 | — | billed on input tokens only (fireworks.ai/pricing prints '$0.1') |
Hardware time
| Model | Price | Note |
|---|---|---|
| H100 80 GB GPUon-demand deployment | $8.00 per hour | Per GPU, billed per GPU second ($0.134/minute); region-restricted 1.5x premium |
| H200 141 GB GPUon-demand deployment | $8.00 per hour | Per GPU, billed per GPU second ($0.134/minute) |
| B200 180 GB GPUon-demand deployment | $13.00 per hour | Per GPU, billed per GPU second ($0.217/minute) |
| B300 288 GB GPUon-demand deployment | $15.00 per hour | Per GPU, billed per GPU second ($0.250/minute) |
| GB300 288 GB GPUon-demand deployment | $20.00 per hour | Per GPU, billed per GPU second ($0.334/minute) |
11 of these 29 prices are ranked in the index; the rest say why in their note. Text and embeddings per 1M tokens unless the row says otherwise.
Discounts and surcharges, as published
- Batch inference is billed at 50% of serverless pricing on both input and output (docs serverless pricing page).
- Prompt-cached input tokens priced lower (per-model cached-input column).
- Bulk / pre-paid discounts by contacting sales: 'We offer discounts for bulk or pre-paid purchases.'
Where these numbers come from
Read on docs.fireworks.ai/serverless/pricing on 16 Oct 2026, checked against 3 more of Fireworks AI’s own pages. Macrostack measures nothing and is paid by no provider on this index. Prices change; the date is the claim.
The other providers in the index
- The AI API price index — every provider, cheapest first, and the download
- OpenAI API pricing — model maker
- Anthropic API pricing — model maker
- Google Gemini API pricing — model maker
- Mistral AI API pricing — model maker
- Cohere API pricing — model maker
- Groq API pricing — open-model host
- Together AI API pricing — open-model host
- Replicate API pricing — open-model host
- Deepgram API pricing — voice
- ElevenLabs API pricing — voice
Questions
- How much does the Fireworks AI API cost?
- On 16 Oct 2026, Fireworks AI's current text models cost $0.20 to $15.00 per million output tokens at the standard tier (NVIDIA Nemotron 3.5 Lightning 30B A3B to Kimi K3). The full list, with input and cached-input prices, is on this page.
- Does Fireworks AI have a free tier?
- What Fireworks AI's own pages say: The pricing page states no free tier or credit amount ('Pay per token with pre-paid, usage-based billing. Simply add a payment method to buy credits'). The official billing FAQ is titled 'What happens when I finish my $1 credit?' and says that without a payment method the account is suspended once that $1 credit is used — i.e. a $1 starting credit is implied, not stated as an offer.
- Does Fireworks AI offer discounts?
- Batch inference is billed at 50% of serverless pricing on both input and output (docs serverless pricing page). Prompt-cached input tokens priced lower (per-model cached-input column). Bulk / pre-paid discounts by contacting sales: 'We offer discounts for bulk or pre-paid purchases.'