What is Pioneer?
Pioneer is an inference API from Fastino Labs that gets smarter as you use it. You send your traffic through Pioneer like any normal inference API, and it automatically spots where your model is underperforming on accuracy, cost, or latency, then trains small specialist models on your own data to close those gaps. The improvement loop runs on its own — you don't write any fine-tuning code or manage GPUs.
Do I have to rewrite my code to use Pioneer?
No. Pioneer is a drop-in replacement for your existing OpenAI or Anthropic client. You point your client's base URL at Pioneer and use a Pioneer API key - your existing code keeps working. There's no migration project.
Which models can I use? Pioneer offers two families. Encoder models (GLiNER2 Large, GLiGuard, GLiNER2-PII) are built for fast, structured tasks like entity extraction, classification, and PII redaction. Decoder models (large language models) include Qwen3, Llama, DeepSeek, Gemma, Nemotron, and Kimi K2.6, plus inference-only access to frontier proprietary models like Claude and GPT through the compatible endpoints. You can see the full live list by calling GET /base-models.
Do I need an ML or MLOps team?
No. Pioneer handles dataset creation, fine-tuning, evaluation, and serving for you. The continuous improvement runs happen automatically and are benchmarked before any traffic is routed to a new model.
Plans, billing & free credits
What plans are available?
There are three: Free ($0/month), Pro ($20/user/month), and Enterprise (custom pricing). Free is best for trying Pioneer and experimenting; Pro suits production workloads that need higher inference caps, downloadable model weights, Deep Research mode, and team members; Enterprise adds private VPC/BYO-cloud deployment, a dedicated GPU fleet, compliance, and an SLA.
What free credits do I get right now? We're running a launch promotion through August. Pro users get $50/day of free inference, and Free users get $30 of credit to try things out. These credits work across any LLM we offer, so you can experiment freely.
What happens when my credits run out or after the promotion ends?
You simply pay for the inference you use. Notably, the model-improvement runs are included — retraining starts at $0, so you're only paying for inference, not for the fine-tuning that makes your models better.
Is there a charge for storing my datasets?
No. Storage is free on every plan — you won't be billed for datasets you upload or create.
API & integration
How do I get an API key? Sign in at pioneer.ai, go to Settings → API Keys, click Create key, name it, and copy the value. You can only see the full key once at creation, so copy it before closing the dialog. If you lose a key, just revoke it and create a new one.
How do I authenticate my requests?
Include your key in the X-API-Key header on every request. There's no OAuth flow and no tokens to refresh. We recommend storing the key in an environment variable or secrets manager rather than hardcoding it, and using a separate key per environment.
Is the API compatible with my existing OpenAI or Anthropic setup?
Yes. Pioneer exposes OpenAI-compatible and Anthropic-compatible endpoints. Set your base URL to https://api.pioneer.ai/v1, swap in your Pioneer key, and your existing SDK calls — including streaming — work unchanged. There's also a native POST /inference endpoint for schema-based extraction and classification.
Is there a CLI? Yes. You can install the Pioneer CLI, authenticate with your API key, and manage datasets, training, and inference from your terminal. Full setup steps are in the docs.
Where can I find error codes and rate limits?
Common HTTP statuses (such as 401 for a missing/invalid key and 402 when credits run low) and current rate limits are documented in the API reference. If you hit a 402, you can add credits or upgrade in Settings → Billing.
Data & privacy
Do you train on my data?
By default, Pioneer may use your data to improve models. You can opt out on Pro and Enterprise plans. Enterprise also lets you run fine-tuning privately inside your own VPC, so your data never leaves your infrastructure.
Who owns my fine-tuned models and datasets? You do. Your model weights and training datasets are yours, and you can download them at any time to bring to another platform or keep training on your own infrastructure.
Is Pioneer compliant with SOC 2 or HIPAA?
Compliance options including SOC 2 and HIPAA, along with private networking and dedicated infrastructure, are available on the Enterprise plan. Contact the team to discuss your requirements.
How do I reach support?
Email [email protected], join the community on Discord, or check live system status on the Pioneer status page.