Tools
Fireworks AI vs Unsloth
Fireworks AI or Unsloth? Their plans monthly and yearly, the capabilities their makers state, security and latest updates, side by side — read from the makers' own pages.
In inglese
In short
- Both: Command line, Choice of models, API
- Only Fireworks AI states: Official SDKs, Runs models for you
- Only Unsloth states: Chat about your code, Runs commands, Self-hosted, Search over your data
Fireworks AI
Fireworks AI hosts, fine-tunes, and serves open models through serverless and dedicated APIs.
Plans
Serverless Inference — Embeddings (up to 150M parameters)
- $0.008 / 1M input tokens
- Per-token pricing
- Zero setup
- No cold starts
- High rate limits
- Postpaid billing
Serverless Inference — Embeddings (150M–350M parameters)
- $0.016 / 1M input tokens
- Per-token pricing
- Zero setup
- No cold starts
- High rate limits
- Postpaid billing
Serverless Inference — Qwen3 8B
- $0.1 / 1M input tokens
- Per-token pricing
- Zero setup
- No cold starts
- High rate limits
- Postpaid billing
Managed Training — Models up to 16B parameters
- $0.50 / 1M training tokens
- $1.00 / 1M training tokens
- $2.00 / 1M training tokens
- Supervised and preference fine-tuning
- Serve fine-tuned models for the same price as base models
Managed Training — Models 16.1B–80B
- $3.00 / 1M training tokens
- $6.00 / 1M training tokens
- $12.00 / 1M training tokens
- Supervised and preference fine-tuning
- Serve fine-tuned models for the same price as base models
Managed Training — Models 80B–300B
- $6.00 / 1M training tokens
- $12.00 / 1M training tokens
- $24.00 / 1M training tokens
- Supervised and preference fine-tuning
- Serve fine-tuned models for the same price as base models
Managed Training — Models over 300B
- $10.00 / 1M training tokens
- $20.00 / 1M training tokens
- $40.00 / 1M training tokens
- Supervised and preference fine-tuning
- Serve fine-tuned models for the same price as base models
Serverless Training API — GLM 5.3
- $4.86 / 1M Prefill
- $0.972 / 1M Cached Prefill
- $12.15 / 1M Sample
- $14.58 / 1M Train
- Shared, always-on trainer pool for LoRA training
- No provisioning or idle cost
- Pay only for tokens prefetched, sampled, and trained
Serverless Training API — Qwen 3.8 27B
- $1.86 / 1M Prefill
- $0.372 / 1M Cached Prefill
- $5.595 / 1M Sample
- $4.103 / 1M Train
- Shared, always-on trainer pool for LoRA training
- No provisioning or idle cost
- Pay only for tokens prefetched, sampled, and trained
Serverless Training API — Kimi K3
- $10.87 / 1M Prefill
- $2.17 / 1M Cached Prefill
- $27.11 / 1M Sample
- $32.55 / 1M Train
- Shared, always-on trainer pool for LoRA training
- No provisioning or idle cost
- Pay only for tokens prefetched, sampled, and trained
Prices checked 2026-09-25 on the maker’s page.
Capabilities
- Command line — “Developers Model Library Docs CLI API Changelog” source
- Choice of models — “Route to the best open or closed model for every task, and cut your AI coding spend 50 to 75%.” source
- API — “Serverless. Pay per token with Priority and Fast options to meet your requirements. OpenAI and Anthropic compatible.” source
- Official SDKs — “The Fireworks Training SDK lets us focus on our research instead of wrestling with infrastructure.” source
- Runs models for you — “Serve the latest open models, or your own trained versions.” source
Security
- SOC 2 Type II — “SOC 2 Type 2” source
- SOC 2 — “SOC 2 Type 2” source
- ISO 27001 — “ISO 27001 Certificate” source
- ISO 42001 — “ISO 42001 Certificate” source
- GDPR — “Compliance SOC 2 Type 2 HIPAA GDPR” source
- HIPAA — “SOC 2 Type 2 HIPAA” source
Latest updates
- Serverless pricing update: DeepSeek V4.1 Flash
Serverless pricing for DeepSeek V4.1 Flash changes; dedicated deployment and Reserved Throughput pricing is unaffected.
- New deployment creation flags: deploymentShape: "default" and acceptShapelessRisk
Create Deployment adds deploymentShape: "default" to pick a validated deployment shape and acceptShapelessRisk to create without a shape.
- Upcoming Serverless deprecation: older DeepSeek, GLM, Muse, and Kimi models
Several older Serverless models will be decommissioned on September 25, 2026; migrate to a recommended replacement before then.
Unsloth
Unsloth runs and fine-tunes AI models locally through a free, open-source desktop app, web UI, and code-based tools.
Plans
Free
Free
- Open-source
- Supports Mistral and Gemma
- Supports Llama 1, 2, and 3
- Supports 4 bit and 16 bit LoRA
unsloth Pro
Contact us
- 2.5x number of GPUs faster than FA2
- 20% less memory than OSS
- Enhanced MultiGPU support
- Up to 8 GPUs support
unsloth Enterprise
Contact us
- 32x number of GPUs faster than FA2
- Up to 30% accuracy
- 5x faster inference
- Full training
- Multi-node support
- Customer support
Prices checked 2026-09-24 on the maker’s page.
Capabilities
- Chat about your code — “Download a model and start chatting in minutes.” source
- Runs commands — “Execute Bash and Python in a secure sandbox so models can run code, test results and complete real tasks locally.” source
- Command line — “then run unsloth start claude .” source
- Choice of models — “Discover, manage and download the right quantization for your device from the built-in model hub.” source
- Self-hosted — “Open-source. Free. 100% Local.” source
- API — “Unsloth also exposes an OpenAI-compatible API, so existing apps, scripts and SDKs can connect to your local models through a familiar interface.” source
- Search over your data — “Private web search, deep research, RAG, MCP + exports (NVFP4, GGUF)” source
Latest updates
- Qwen-Image-2.1 + Skills (Qwen-Image-2.1)
Adds local Qwen-Image-2.1, custom Agent Skills, draggable chats, faster reasoning blocks, and Linux update and installation options.
- Qwen-Image-2.1 + Skills (Qwen-Image-2.1)
Adds local Qwen-Image-2.1, custom Agent Skills, draggable chats, faster reasoning blocks, and Linux update and installation options.
- Docker + Multi User + AMD Support (v0.1.810-beta)
Adds a Docker image with NVIDIA and AMD support, multi-user accounts, diffusion support, ARM64 Windows CUDA support, and RDNA1/2 support.