Provider catalog
The lab aggregates a deliberately multi-vendor catalog: no single provider
is a dependency, and every tier has at least one alternative. Providers are
grouped by how they are accessed.
LLM providers
Direct APIs
| Provider |
Models in the catalog |
Notes |
| DeepSeek |
V4.1 Flash, V4 Pro |
Direct legs of the fast chains |
| MiniMax |
M3.1 Flash |
Free tier (chat-free) |
| Cerebras, Groq |
Qwen 3.8 27B |
Very fast inference (chat-ultra-fast) and free legs |
Plans and subscription shims (OpenAI-compatible proxies over plans)
| Plan |
Exposed as |
Pattern |
| Alibaba Model Studio token plan |
chat-fast, chat-balanced |
Fixed weekly credit plan; chains fall back to PAYG when it pauses |
| ChatGPT plan |
sub/openai/… |
Plan proxy holds plan credentials, refreshes tokens, serves /v1 to LiteLLM |
| Claude and Gemini plans |
sub/anthropic/…, sub/google/…, chat-smart, vision |
CLIProxyAPI shim over the plan credentials |
Aggregators (pay-as-you-go)
| Aggregator |
Models via catalog |
Role |
| Nous Research |
DeepSeek V4.1 Flash, MiMo v2.6 Pro/Flash, Gemini 3.8 Flash, LongCat (free) |
Main PAYG aggregator: backs research-deep and the coding tiers |
| RunAnywhere |
DeepSeek V4.1 Flash, MiMo v2.6 Flash, GLM-5.3 Flash |
Fastest endpoint; backs chat-main |
| TokenRouter |
Qwen 3.8 Max, GLM-5.3 |
PAYG legs for diversity |
| OpenRouter |
One free model |
Free fallback leg only |
| OpenAI (direct) |
Vision fallback, embeddings |
Separate credit pool for special-purpose models |
Capability providers (via the Provider Broker)
| Capability |
Providers |
Failover chain |
| Web search |
Nous Fast Search, Tavily, SearXNG (self-hosted), Serper, Brave |
nous-fast-search → tavily → searxng → serper → brave |
| Research |
Perplexity |
sonar / sonar-pro only |
| Scraping |
Crawl4AI (self-hosted), Firecrawl, Nous Firecrawl |
crawl4ai → firecrawl → nous-firecrawl |
| Speech-to-text |
ElevenLabs, Deepgram, OpenAI Whisper |
elevenlabs → deepgram → openai |
| Text-to-speech |
Deepgram, OpenAI, ElevenLabs |
deepgram → openai (ElevenLabs first for selected agents) |
| Voice calls |
Telnyx |
Signed webhooks, call control |
| Browser automation |
Self-hosted browser service |
Self-hosted only (cloud browser forbidden) |
| Image generation |
Higgsfield |
Degraded-capability reporting |
| MCP services |
Scrapling and other registered servers |
Host allowlisted |
| Agent integrations |
Composio |
Least-privilege fixed operations |
Why so many providers?
- Resilience: a plan outage becomes a routing change, not an outage. The
platform regularly lives through provider incidents without users noticing.
- Cost: the same workload can be served at wildly different prices; tiers
and mirrors let policy pick the cheapest rung that satisfies the
requirement.
- Evaluation: running the same agent across vendors produces real
comparative data — which model is actually better for this workload is
decided by measurement, not marketing.
Sourcing discipline
- Each provider is pinned in the alias catalog with its exact endpoint, key
source, and documented quirks (e.g., endpoint differences between a plan
API and a general API; user-agent gates; output-limit mapping).
- Subscription-backed aliases are explicitly routed through the outer agent
chain's fallback order — never through invisible proxy-level fallbacks — so
the failover behavior is deterministic and reviewable.
- Keys live in Infisical and are projected into the proxy and broker
deployments only.