Pricing | Fireworks (https://fireworks.ai/pricing) citeturn140view1 [wordlim: 200] Crawled: today; Content type: text/html; Source: open({"ref_id":"https://fireworks.ai/pricing","lineno":null}); Total lines: 410 L0: 【0†The right model for every task. Try Nexus†nexus.fireworks.ai】 L1: L2: 【1†Docs†docs.fireworks.ai】 L3: L4: 【2†Log in†app.fireworks.ai】 L5: L6: * 【3†Nexus†nexus.fireworks.ai】 L7: * 【2†Platform†app.fireworks.ai】 L8: L9: 【4†Get started】 L10: L11: * 【5†Nexus For work and co-work†nexus.fireworks.ai】 L12: * 【6†Platform For systems and apps†app.fireworks.ai】 L13: L14: Pricing L15: # Pricing that scales L16: from idea to enterprise. L17: L18: Start on pay-per-token, reserve throughput as you scale, or run Fireworks inside your own cloud. Explore Platform for systems and apps or Nexus for code and co-work. L19: L20: 【7†Get started†app.fireworks.ai】【8†Talk to an expert】 L21: L22: Platform Nexus L23: L24: Inference Training L25: L26: ## Deploy on your terms. L27: L28: Start on shared serverless capacity, guarantee it as traffic grows, or run Fireworks inside your cloud. L29: ### Serverless L30: L31: Pay per token · From $0.05 / 1M input tokens L32: L33: Call any serverless model in seconds. No commitment. Usage draws down prepaid credits. L34: L35: Best for: Launching fast and spiky traffic L36: L37: * Standard, Priority, Fast, Batch and Spot modes L38: * Cached input billed at half price or less, with no write fee L39: * Monthly spend limits and Auto Reload L40: L41: Commitment: No L42: L43: 【7†Get started†app.fireworks.ai】 L44: ### Reserved Throughput L45: L46: Reserve dollars per minute · Serverless rates, reserved by the minute L47: L48: Guaranteed throughput on serverless and on-demand models, backed by an SLA. L49: L50: Best for: Steady production traffic L51: L52: * Throughput SLA and 99.9% uptime terms L53: * Adds to your adaptive rate limits L54: * Usage above your reservation can bill at standard Serverless rates, never a penalty rate L55: L56: Commitment: Yes L57: L58: 【8†Talk to us】 L59: ### Fireworks Sovereign L60: L61: Contract · Talk to us L62: L63: Private preview for Enterprise L64: L65: Fireworks-managed inference on GPUs in your cloud account or data center. L66: L67: Best for: Data residency and existing GPU capacity L68: L69: * Inference and response handling stay inside your boundary L70: * Achieve over 2x tokens per GPU dollar with our engine L71: * Optional overflow to Fireworks-managed capacity L72: L73: Commitment: Yes L74: L75: 【8†Talk to us】 L76: ## Pick a model. Pick a mode. L77: L78: Every serverless request bills on input, cached input and output tokens. Switch modes per request, on the same endpoint. L79: L80: ### Standard L81: L82: The default for most traffic. L83: L84: ### Priority L85: L86: Prioritize your traffic during peak load times. L87: L88: ### Fast L89: L90: 100+ tokens per second for interactive apps. L91: L92: ### Ultra-fast L93: L94: New: Available for specific models, in preview. Talk to us. L95: L96: ### Batch L97: L98: 50% off input and output for jobs that don’t need an immediate answer. L99: ### Spot L100: L101: 30% off Standard on best-effort capacity, available for select models only. Throughput can drop to zero, so retry or fall back to Standard. L102: L103: ### Highlighted Serverless rates: Standard and Priority L104: ## L105: L106: USD per 1M tokens. L107: Highlighted Serverless rates, Standard and Priority (USD per 1M tokens) L108: Model | Input | Cached input | Output | Priority (input / cached / output) L109: --- | --- | --- | --- | --- L110: Ember-1 | $3.00 | $0.30 | $15.00 | $3.75 / $0.375 / $18.75 L111: Kimi K3 | $3.00 | $0.30 | $15.00 | $3.75 / $0.375 / $18.75 L112: Kimi K3.1 Pro | $2.00 | $0.20 | $10.00 | $3.00 / $0.30 / $15.00 L113: Qwen 3.8 Max | $2.00 | $0.25 | $6.00 | $3.00 / $0.375 / $9.00 L114: GLM 5.3 | $1.40 | $0.26 | $4.40 | $1.75 / $0.325 / $5.50 L115: Muse Spark Open | $1.25 | $0.15 | $4.25 | $1.875 / $0.225 / $6.375 L116: NVIDIA Nemotron 3 Ultra (Preview) | $0.60 | $0.12 | $2.40 | $0.75 / $0.15 / $3.00 L117: DeepSeek V4.1 Flash | $0.30 | $0.006 | $1.20 | $0.375 / $0.0075 / $1.50 L118: DeepSeek V4.1 Flash 1017 | $0.30 | $0.006 | $1.20 | $0.375 / $0.0075 / $1.50 L119: MiniMax M3 | $0.30 | $0.06 | $1.20 | $0.45 / $0.09 / $1.80 L120: GLM 5.3 Flash | $0.15 | $0.03 | $0.50 | $0.1875 / $0.0375 / $0.625 L121: OpenAI GPT OSS 120B | $0.15 | $0.015 | $0.60 | $0.18 / $0.018 / $0.72 L122: NVIDIA Nemotron 3.5 Lightning 30B A3B | $0.05 | $0.01 | $0.20 | $0.0625 / $0.0125 / $0.25 L123: ### Spot rates L124: ## L125: L126: USD per 1M tokens. L127: L128: US-only Serverless is available on select models at a 50% premium (waived on Muse Spark Open); 【9†see pricing here†docs.fireworks.ai】. Models not listed here are priced by size. L129: L130: 【10†See all models and rates†docs.fireworks.ai】 L131: Spot rates (USD per 1M tokens) L132: Model | Tier | Cached input | Uncached input | Output L133: --- | --- | --- | --- | --- L134: GLM-5.3 | Standard | $0.26 | $1.40 | $4.40 L135: GLM-5.3 | Spot | $0.182 | $0.98 | $3.08 L136: Kimi K3 | Standard | $0.30 | $3.00 | $15.00 L137: Kimi K3 | Spot | $0.21 | $2.10 | $10.50 L138: DeepSeek V4.1 Flash | Standard | $0.006 | $0.30 | $1.20 L139: DeepSeek V4.1 Flash | Spot | $0.0042 | $0.21 | $0.84 L140: DeepSeek V4.1 Flash 1017 | Standard | $0.006 | $0.30 | $1.20 L141: DeepSeek V4.1 Flash 1017 | Spot | $0.0042 | $0.21 | $0.84 L142: ### Embeddings L143: L144: Embeddings (USD per 1M input tokens) L145: Base model | Price L146: --- | --- L147: Up to 150M parameters | $0.008 L148: 150M to 350M parameters | $0.016 L149: Qwen3 8B | $0.10 L150: L151: ## Lower the cost of every finished task. L152: L153: Token price is one input. Cache hits, the mode you choose and the model each turn goes to decide the rest. L154: L155: ### Keep context cached L156: L157: Session affinity keeps each agent on the server that already holds its context. Long runs reach cache hit rates up to 99.6%. L158: L159: 【11†Prompt caching†docs.fireworks.ai】 L160: ### Batch what can wait L161: L162: Async jobs run at half the Standard rate. L163: L164: 【12†Batch inference†docs.fireworks.ai】 L165: L166: ### Send background work to Spot L167: L168: Take 30% off for traffic that can retry or fall back L169: L170: 【13†Spot Serverless†docs.fireworks.ai】 L171: L172: ### Route each turn L173: L174: Send routine work to open models and save premium models for the turns that earn it. L175: ## Pick your training path, from co-development with our proven researchers to running your own loop. L176: L177: Train at the rates below, then serve a fine-tuned model on Serverless or a Dedicated Deployment. See our inference rates or talk to us about a custom deployment. L178: L179: ## Muse Glimmer 30B SFT is free through November 30 L180: L181: In the UI and the Serverless Training API. No credit card required. L182: L183: 【14†Start a job†app.fireworks.ai】 L184: ### Serverless Training API L185: L186: By model · see table below L187: L188: Customize your own training loop L189: L190: Best for: ML engineers and ML researchers L191: L192: * Write your loop in Python and attach to an always-on trainer pool L193: * Nothing to provision, no idle cost. You pay for the tokens you prefill, sample and train L194: * LoRA on popular models, including RL L195: L196: 【15†Start training†docs.fireworks.ai】 L197: ### Dedicated Training API L198: L199: From $8.00 / GPU-hour L200: L201: You run the loop, at full scale L202: L203: Best for: Advanced research teams L204: L205: * Provisioned trainers and deployments L206: * Full-parameter training, the largest MoE models, longer context and higher LoRA ranks L207: * Billed at On Demand GPU rates L208: L209: 【8†Talk to us】 L210: L211: We build with you L212: L213: ## Forward-deployed researchers and engineers train your model with your team, from eval design through production. L214: L215: Priced per engagement. L216: L217: 【8†Talk to Fireworks Lab】 L218: ### Serverless Training API L219: ## L220: L221: Billed per 1M tokens you prefill, sample and train. Repeated prefixes bill 80% off. L222: Serverless Training API (USD per 1M tokens) L223: Base model | Context | Prefill | Cached prefill | Sample | Train L224: --- | --- | --- | --- | --- | --- L225: Kimi K3 | 192K | $10.87 | $2.17 | $27.11 | $32.55 L226: Muse Spark Open | 256K | $6.19 | $1.239 | $15.48 | $18.58