Four bits to a deployed model.
Record · train · eval · deploy.
Fourbit is the managed fine-tuning platform for LLMs, VLMs, VLAs, and diffusion models. We own all four stages so you don't have to — your data goes in, a working deployed model comes out.
Every modality. One workflow.
Pricing is on every card. Pick a base, point at your data, and we wire up the right trainer.
Llama 3.3 70B
Meta · 128k ctx
Frontier open-weights generalist. Best-in-class for instruction-tuning at the 70B tier.
Qwen3 72B
Alibaba · 256k ctx
Strong multilingual generalist with long context. Excellent for non-English and code tasks.
Qwen2.5-VL 72B
Alibaba · Image · video · OCR
Frontier open VLM. Great for chart, table, and document understanding.
Pi-0
Physical Intelligence · Bimanual · 50Hz
Generalist VLA from Physical Intelligence. Bimanual manipulation out of the box.
FLUX.1 [dev]
Black Forest Labs · Up to 2048×2048
Frontier text-to-image. The default for character, product, and brand LoRAs.
Launch a fine-tune in four lines.
Use the CLI for one-offs, the SDK for pipelines, or the dashboard when you want to click. Same primitives, same job IDs across all three.
- Resumes from the latest checkpoint on preemption.
- Streams loss, grad-norm, and sample generations in real time.
- Promotes a checkpoint to a private endpoint with one call.
$ fourbit train \
--base qwen3-8b \
--data s3://acme/support-tickets.jsonl \
--recipe lora \
--budget 40usd
→ job ftjob_8r2hQv queued · 4×H100 · eta 38m
→ stream: https://fourbit.ai/r/ftjob_8r2hQvFour bits to a deployed model.
Record, train, eval, deploy. Every fine-tune in fourbit moves through these four stages — and we own all of them so you don't have to.
Record
Bring your data. We handle the rest.
Upload teleop episodes, JSONL, image sets, or point us at a bucket. We tokenize, shard, and pack — and tell you up-front if your data is the bottleneck.
Train
Curated recipes per architecture. Per-second billing.
LoRA, QLoRA, full FT, DPO. The right hardware lands automatically. Resumes from checkpoint on preemption. Stream loss live.
Eval
Real numbers, not vibes.
Sim eval, task-specific benchmarks, side-by-side with the base model. The scorecard your boss can read — before anything ships to production.
Deploy
Tuned weights are not a product. A working endpoint is.
One click to a private inference endpoint, or pull weights for your stack. 30-day on-call engineer included with every pilot.
The boring parts, handled.
Recipes that just work
Curated training configs per architecture. No more chasing flash-attn versions or DeepSpeed YAML.
Eval that means something
Run task-specific evals on every checkpoint. Compare runs side-by-side, promote winners.
Bring your hardware
Run on our pooled H100/H200 capacity, or attach your own cluster. Same dashboard either way.
Pay for what you train
Per-second GPU billing, transparent pricing. Cap your spend before you queue.
Train your first model today.
Free credits to get your first run on the dashboard. No card needed.