Needle pricing
Generate synthetic training examples, train on hosted GPUs, compare accuracy across model sizes, and download models to run locally. No GPU setup required; downloaded models remain usable after your plan ends.
- Starter
- Starter: $19 one-time for 3 fine-tunes, 3,000 generated examples, and 10 GB storage. Use within 30 days. Once per account before subscribing; no automatic renewal.
- Standard
- Standard: $99 per month for 10 fine-tunes, 10,000 generated examples, and 10 GB storage. Run and generation allowances reset each billing period. Cancel anytime.
- Custom
- Custom: contact us for higher limits.
Try the free base model first. Choose Starter to test a use case and iterate before committing; choose Standard for ongoing development. If the base model already meets your requirements, you do not need to pay.
Choose a plan- Starter is a one-time purchase for one account before subscribing. Access and unused allowances expire after 30 days; there is no automatic renewal.
- Standard is a monthly subscription. Upgrading starts fresh Standard allowances and preserves files and models. Unused Starter allowances do not carry over, its payment is not credited, and it cannot be purchased again.
- Both plans include the same model sizes, hosted GPU training, synthetic example generation, validation and test scores, model downloads, and playground access. Training and generation spend separate allowances; generated datasets can be reused across runs.
- Storage includes uploaded files, generated datasets, model exports, and reserved output space. Jobs reserve up to 512 MiB for output. Failed jobs return their allowance; cancelled jobs do not.
- After access ends, files and models remain available to download for another 30 days before they may be deleted. An active paid plan preserves them. Downloaded models remain usable locally.
- Generation creates synthetic examples from tool definitions and optional product context. Uploaded datasets are not automatically cleaned or augmented. Review generated data and evaluate on independent examples from your use case.
- There is no hosted inference endpoint. Run downloaded models locally with the open-source runtime. Accuracy improvements depend on your data and task.