BTL Runtime
LLM 7.2/10OpenAI-compatible inference gateway from Bad Theory Labs with a $5 signup credit and 346 public model slugs. Its public btl-2 slug currently maps to one Google AI Studio Gemma route rather than a published multi-provider auto-router.
BTL Runtime scores 7.2/10 for free value in LLM. The main constraint: $5 signup credit, no card required. The public model catalog lists 346 public slugs across 9 routed providers; btl-2 is positioned as the first test route for stretching the starter credit and maps to google-ai-studio / gemma-4-31b-it.
Free value score
7.2 / 10Show the math
Key metrics
Models
7 free · 346+ total| Model | Family | AA | Access |
|---|---|---|---|
| Claude Opus 4.8 | Claude | 55.7 | free |
| GPT-5.5 | GPT | 54.8 | free |
| Gemini 3.1 Pro | Gemini | 46.5 | free |
| DeepSeek V4 Pro | DeepSeek | 44.3 | free |
| Kimi K2.7 Code | Kimi | 42 | free |
| btl-2 | Gemma | — | free |
| Llama 3.3 70B Instruct | Llama | — | free |
Showing first 7 of a larger catalog.
What you get
$5 signup credit with no card required
OpenAI-compatible base URL at https://api.badtheorylabs.com/v1
346 public model slugs across 9 routed providers
btl-2 maps to google-ai-studio / gemma-4-31b-it, Gemini protocol, text I/O, streaming, 262K context, and 8K max output
btl-2 route benchmark price is $0/M input/output; sell price is shown as $5/M input/output/cache-read
Per-request savings headers, usage ledger, routing, cache, and dedupe
Limitations
- Finite signup credit
The public pricing page confirms a $5 signup credit; no public page confirms a renewable monthly token allowance.
- btl-2 auto-router claim unverified
Public catalog data shows btl-2 as one google-ai-studio route to gemma-4-31b-it, not a GLM/OpenAI/Anthropic/Gemma auto-router.
Sources
-
"$5 free on signup. No card required. You see every provider cost."
https://runtime.badtheorylabs.com/ -
"Point the same OpenAI-compatible client at https://api.badtheorylabs.com/v1."
https://runtime.badtheorylabs.com/docs -
"Catalog snapshot 346 public slugs across 9 providers."
https://runtime.badtheorylabs.com/models -
"btl-2. Best first test if you want the starter credit to last as long as possible."
https://runtime.badtheorylabs.com/models -
"btl-2 is available through BTL Runtime across 1 routes and 1 providers."
https://runtime.badtheorylabs.com/models/btl-2 -
"Providers google-ai-studio. Protocol families gemini. Input modalities text. Output modalities text."
https://runtime.badtheorylabs.com/models/btl-2 -
"Provider model id gemma-4-31b-it."
https://runtime.badtheorylabs.com/models/btl-2 -
"Benchmark input $0.00 / 1M. Benchmark output $0.00 / 1M."
https://runtime.badtheorylabs.com/models/btl-2 -
"Each response includes the direct provider cost, what you were charged, and the amount saved."
https://runtime.badtheorylabs.com/