# Fastino Labs

**Canonical:** https://apis.io/providers/fastino-labs/  
**Website:** https://www.fastino.ai/  
**APIs profiled:** 1

Fastino Labs is an applied research lab building production-ready small language models (SLMs) for agentic AI — structured data extraction, classification, PII detection, and safety moderation. Its Pioneer platform is an agentic fine-tuning and inference API that lets teams generate synthetic training data, fine-tune GLiNER encoder models and open-source decoder LLMs (Qwen, Llama, DeepSeek, Gemma), evaluate them against labeled datasets, and serve inference through native, OpenAI-compatible, and Anthropic-compatible endpoints. Fastino also maintains the open-source GLiNER2 schema-based information-extraction model and the GLiGuard safety-moderation model.

## Kin Score — 48.4 / 100 (developing)

Scored 2026-08-30 under rubric 0.17.2. Trend: flat (-0.6 from 49.0).

| Facet | Score |
|---|---|
| Discoverability | 75.9 |
| Contract Quality | 61.1 |
| Governance | 4.5 |
| Contract Governance | 4.5 |
| Operational Transparency | 55.3 |
| Developer Ergonomics | 78.6 |
| Commercial Clarity | 10.5 |
| Access Clarity | 10.5 |

## Agent readiness — 52.0 (agent-ready)

| Dimension | Value |
|---|---|
| Spec Presence | yes |
| Agentic Access | derived |
| Reversibility Documented | no |
| MCP Server | documented |
| Auth Clarity | served |
| Idempotency | no |
| Error Semantics | documented |
| OpenAPI Examples | verified |
| Rate Limit Signal | verified |
| Event Surface Described | no |
| Agent Skills | yes |
| Well Known Catalog | yes |
| Consent Identity | no |
| Agent Card | no |
| Dry Run Mode | no |
| Delegated Identity | served |
| Protected Resource Metadata | verified |
| Dynamic Client Registration | no |
| Agentic Commerce | no |

## Access

Self-serve signup — onboarding: self-serve, pricing: unknown, trial: no (confidence: medium).

## APIs (4)

- **Fastino Labs anthropic-compat API** — Anthropic-compatible endpoints. Use with the Anthropic SDK by setting base_url=https://api.pioneer.ai.
- **Fastino Labs inference API** — Pioneer-native inference endpoint (encoder NER/classification/extraction and decoder text generation).
- **Fastino Labs inference-history API** — List and retrieve past inference records.
- **Fastino Labs openai-compat API** — OpenAI-compatible endpoints. Use with the OpenAI SDK by setting base_url=https://api.pioneer.ai/v1.

## MCP servers (1)

- **Fastino Labs MCP Server** — Mintlify-hosted documentation MCP server for the Pioneer docs.

## Agentic access (1)

- **Fastino Labs Agentic Access** — 8 operations · 5 acting

## Security (3)

- **Fastino Labs Authentication** — apiKey/http · 2 schemes
- **Fastino Labs Domain Security** — TLSv1.3 · HSTS · DMARC
- **Fastino Labs Vulnerability Disclosure** — security.txt · contact published

## Tags

Company, Artificial Intelligence, Machine-Learning, Small Language Models, Fine-Tuning, Inference, Named Entity Recognition, Information Extraction, LLM, Agents, PII Detection, Model Training

---

Profiled by [API Evangelist](https://apievangelist.com) and published on [APIs.io](https://apis.io/providers/fastino-labs/). Scores are computed from the provider's own public artifacts under a published rubric.