# Fireworks AI pricing Extracted automatically from [https://fireworks.ai/pricing](https://fireworks.ai/pricing), captured 2026-10-17 (extracted 2026-10-18). Only prices whose exact wording was found on the page under the right plan are listed; always confirm on the vendor’s site. - Pricing model: Usage-based - Starts at: not listed - Badges: Usage-based - Confidence: Medium: some prices failed a check and are hidden - Company profile: [Fireworks AI](https://rewindex.com/company/fireworks-ai.md) · Pricing page changes: [history](https://rewindex.com/company/fireworks-ai/changes.md?kind=pricing) ## Plans | Plan | Product | Price | Source quote | | --- | --- | --- | --- | | On demand deployments | On-Demand Pricing | $0.134 per minute (H100 80 GB GPU) | “H100 80 GB GPU $0.134 $8.00” | | | | $8 per hour (H100 80 GB GPU) | “H100 80 GB GPU $0.134 $8.00” | | | | $0.134 per minute (H200 141 GB GPU) | “H200 141 GB GPU $0.134 $8.00” | | | | $8 per hour (H200 141 GB GPU) | “H200 141 GB GPU $0.134 $8.00” | | | | $0.217 per minute (B200 180 GB GPU) | “B200 180 GB GPU $0.217 $13.00” | | | | $13 per hour (B200 180 GB GPU) | “B200 180 GB GPU $0.217 $13.00” | | | | $0.25 per minute (B300 288 GB GPU) | “B300 288 GB GPU $0.250 $15.00” | | | | $15 per hour (B300 288 GB GPU) | “B300 288 GB GPU $0.250 $15.00” | | | | $0.334 per minute (GB300 288 GB GPU) | “GB300 288 GB GPU $0.334 $20.00” | | | | $20 per hour (GB300 288 GB GPU) | “GB300 288 GB GPU $0.334 $20.00” | | Embeddings | Serverless Inference | $0.008 per 1,000,000 1M tokens (up to 150M) | “up to 150M $0.008 150M - 350M $0.016” | | | | $0.016 per 1,000,000 1M tokens (150M - 350M) | “up to 150M $0.008 150M - 350M $0.016” | | | | $0.1 per 1,000,000 1M tokens (Qwen3 8B) | “150M - 350M $0.016 Qwen3 8B $0.1” | | DeepSeek V4.1 Flash | Serverless Training API | $0.6 per 1,000,000 1M tokens (Prefill) | “DeepSeek V4.1 Flash 256K $0.60 $0.12 $3.60 $4.98” | | | | $0.12 per 1,000,000 1M tokens (Cached Prefill) | “DeepSeek V4.1 Flash 256K $0.60 $0.12 $3.60 $4.98” | | | | $3.6 per 1,000,000 1M tokens (Sample) | “DeepSeek V4.1 Flash 256K $0.60 $0.12 $3.60 $4.98” | | | | $4.98 per 1,000,000 1M tokens (Train) | “DeepSeek V4.1 Flash 256K $0.60 $0.12 $3.60 $4.98” | | DeepSeek V4 Flash 0731 | Serverless Training API | $1.74 per 1,000,000 1M tokens (Prefill) | “DeepSeek V4 Flash 0731 262K $1.74 $0.35 $4.33 $5.20” | | | | $0.35 per 1,000,000 1M tokens (Cached Prefill) | “DeepSeek V4 Flash 0731 262K $1.74 $0.35 $4.33 $5.20” | | | | $4.33 per 1,000,000 1M tokens (Sample) | “DeepSeek V4 Flash 0731 262K $1.74 $0.35 $4.33 $5.20” | | | | $5.2 per 1,000,000 1M tokens (Train) | “DeepSeek V4 Flash 0731 262K $1.74 $0.35 $4.33 $5.20” | | GLM 5.3 | Serverless Training API | $4.86 per 1,000,000 1M tokens (Prefill) | “GLM 5.3 262K $4.86 $0.972 $12.15 $14.58” | | | | $0.972 per 1,000,000 1M tokens (Cached Prefill) | “GLM 5.3 262K $4.86 $0.972 $12.15 $14.58” | | | | $12.15 per 1,000,000 1M tokens (Sample) | “GLM 5.3 262K $4.86 $0.972 $12.15 $14.58” | | | | $14.58 per 1,000,000 1M tokens (Train) | “GLM 5.3 262K $4.86 $0.972 $12.15 $14.58” | | GLM 5.3 Flash | Serverless Training API | $2.96 per 1,000,000 1M tokens (Prefill) | “GLM 5.3 Flash 200K $2.96 $0.593 $7.41 $8.89” | | | | $0.593 per 1,000,000 1M tokens (Cached Prefill) | “GLM 5.3 Flash 200K $2.96 $0.593 $7.41 $8.89” | | | | $7.41 per 1,000,000 1M tokens (Sample) | “GLM 5.3 Flash 200K $2.96 $0.593 $7.41 $8.89” | | | | $8.89 per 1,000,000 1M tokens (Train) | “GLM 5.3 Flash 200K $2.96 $0.593 $7.41 $8.89” | | Kimi K3 | Serverless Training API | $10.87 per 1,000,000 1M tokens (Prefill) | “Kimi K3 192K $10.87 $2.17 $27.11 $32.55” | | | | $2.17 per 1,000,000 1M tokens (Cached Prefill) | “Kimi K3 192K $10.87 $2.17 $27.11 $32.55” | | | | $27.11 per 1,000,000 1M tokens (Sample) | “Kimi K3 192K $10.87 $2.17 $27.11 $32.55” | | | | $32.55 per 1,000,000 1M tokens (Train) | “Kimi K3 192K $10.87 $2.17 $27.11 $32.55” | | Muse Glimmer 30B | Serverless Training API | $1.96 per 1,000,000 1M tokens (Prefill) | “Muse Glimmer 30B 128K $1.96 $0.39 $4.88 $5.86” | | | | $0.39 per 1,000,000 1M tokens (Cached Prefill) | “Muse Glimmer 30B 128K $1.96 $0.39 $4.88 $5.86” | | | | $4.88 per 1,000,000 1M tokens (Sample) | “Muse Glimmer 30B 128K $1.96 $0.39 $4.88 $5.86” | | | | $5.86 per 1,000,000 1M tokens (Train) | “Muse Glimmer 30B 128K $1.96 $0.39 $4.88 $5.86” | | Muse Spark Open | Serverless Training API | $6.19 per 1,000,000 1M tokens (Prefill) | “Muse Spark Open 256K $6.19 $1.239 $15.48 $18.58” | | | | $1.24 per 1,000,000 1M tokens (Cached Prefill) | “Muse Spark Open 256K $6.19 $1.239 $15.48 $18.58” | | | | $15.48 per 1,000,000 1M tokens (Sample) | “Muse Spark Open 256K $6.19 $1.239 $15.48 $18.58” | | | | $18.58 per 1,000,000 1M tokens (Train) | “Muse Spark Open 256K $6.19 $1.239 $15.48 $18.58” | | Qwen 3.8 27B | Serverless Training API | $1.86 per 1,000,000 1M tokens (Prefill) | “Qwen 3.8 27B 128K $1.86 $0.372 $5.595 $4.103” | | | | $0.372 per 1,000,000 1M tokens (Cached Prefill) | “Qwen 3.8 27B 128K $1.86 $0.372 $5.595 $4.103” | | | | $5.6 per 1,000,000 1M tokens (Sample) | “Qwen 3.8 27B 128K $1.86 $0.372 $5.595 $4.103” | | | | $4.1 per 1,000,000 1M tokens (Train) | “Qwen 3.8 27B 128K $1.86 $0.372 $5.595 $4.103” | | Managed Training | Training | $0.5 per 1,000,000 1M tokens (Models up to 16B parameters — LoRA SFT) | “Models up to 16B parameters $0.50 $1.00 $1.00 $2.00” | | | | $1 per 1,000,000 1M tokens (Models up to 16B parameters — LoRA DPO) | “Models up to 16B parameters $0.50 $1.00 $1.00 $2.00” | | | | $1 per 1,000,000 1M tokens (Models up to 16B parameters — Full Param SFT) | “Models up to 16B parameters $0.50 $1.00 $1.00 $2.00” | | | | $2 per 1,000,000 1M tokens (Models up to 16B parameters — Full Param DPO) | “Models up to 16B parameters $0.50 $1.00 $1.00 $2.00” | --- Canonical: https://rewindex.com/company/fireworks-ai/pricing · Markdown: https://rewindex.com/company/fireworks-ai/pricing.md · JSON: https://rewindex.com/api/v1/companies/fireworks-ai/pricing · Generated 2026-10-18T06:42:00.173Z