MCP & Integrations
TokenLab
Last updated 2026-08-16 · benchmark measured 2026-10-23 — deterministic & reproducible
TokenLab MCP lets AI agents call models and multimodal tools via a unified API.
TokenLab at a glance
TokenLab MCP is a free, open-source connector that lets AI clients like Claude Desktop, Codex, and Cursor call TokenLab's unified API for text, image, video, music, 3D, and audio generation, plus model search and price comparison, for AI agents and developers.
Good for
- Searching and comparing prices across AI models without needing an API key
- Letting an AI agent chat and generate images, video, music, 3D, or audio through one MCP connection
- Expanding to the full tool set for batch requests, the Responses API, world generation, and asset library handling
Pricing
MCP service is free and open source. API calls use prepaid credits, no subscription: top up $5, $10, $20 (+$5 first-time bonus), $50, or $100.
Questions people ask
Do I need an API key to use TokenLab MCP?
No key is needed for the catalog profile, which only searches models and checks pricing. The core and full profiles need a TOKENLAB_API_KEY for chatting, generating media, and handling files or tasks.
Which tool profile should I pick: catalog, core, or full?
Use catalog if you only want to look up models, core (the default, 32 tools) if you want the agent to generate content, and full (89 tools) if you need batch requests, the Responses API, world generation, or asset library features.
How much does TokenLab MCP cost to use?
The MCP service itself is free and open source; any tool call that invokes a model is billed at TokenLab's normal per-model pricing from your prepaid balance.
Does my TokenLab balance expire?
No, the balance never expires and stays in your account until you spend it.
Do I also need to install TokenLab Skills?
No, MCP works on its own; Skills adds guidance such as which model to pick and how to check results, and uses the MCP tools if both are installed.
Written by Legit.Show from tokenlab.sh on 2026-10-19. It does not change the score.
✓ Verified by the maker on 2026-10-17. Ownership was proven with a meta tag, a DNS record or GitHub admin rights. It does not change the score.
Launched something? Add your product, free.
To cite this score: legit.show/s/tokenlab-sh/2026-10-23. That address never changes; this page moves with every re-measure.
Legit.Show scored TokenLab 96/100 on 2026-10-23, measured across all 7 frames from its public surface. legit.show/s/tokenlab-sh/2026-10-23
Is TokenLab production-ready?
Legit.Show scores TokenLab 96 out of 100 — the simple average of its 7 measured frames. Legit.Show ran its deterministic 7-Frame production-readiness benchmark on TokenLab (public-surface assessment), measured from the public surface with no LLM in the scoring path. Its strongest frame is Security; its weakest is Performance. Every frame it averages is published with its evidence on the Legit.Show listing.
The 7 Frames
- Performance — 74/100
- Accessibility — 97/100
- Security — 100/100How this compares across the catalogue: The Web Security Baseline
- Privacy — 100/100How this compares across the catalogue: The Privacy Gap
- Reliability — 100/100
- Standards — 100/100
- Discoverability — 100/100
What we measured
- Security headers present: CSP, HSTS, X-Frame-Options, X-Content-Type-Options, Referrer-Policy.
- Served over HTTPS with a valid certificate.
- Real Lighthouse performance run — 292 ms to first byte.
- Returns a proper 404 for unknown routes.
- 3 of 3 sampled routes reachable.
- Has a reachable privacy policy.
- Sets cookies / loads scripts with no consent prompt.
- Discoverable: structured data, sitemap, OpenGraph image, canonical URL.
Who it's for
AI agents · Claude users · Developers · LLM integrations · MCP clients
Sources and updates
- Description
- Taken from tokenlab.sh's own website on 2026-08-16.
- Operator
- TOKENLAB AI INC, as named in tokenlab.sh's terms, privacy page or footer.
- Benchmark
- Measured by Legit.Show from the live site, the way any visitor sees it — with no access to its code or accounts. Same method for every product, and no AI decides the score. Last checked 2026-10-23.
Put together from public information. TokenLab's maker verified ownership on 2026-10-17, which does not change the score. If anything here is wrong, tell us and a person will check it.
Visit TokenLab → · Alternatives to TokenLab → · How this was measured →