AIWorkbench.dev — Test, Compare & Debug AI Models

//the open tab for ai engineers

Test, compare and debug Claude · GPT-4o · Gemini · Bedrock in one place. Bring your own key — never touches our servers.

8core tools

4providers

$0to start

OSSopen source

BYOK▶ open workbench

[01]Core Tools

4 tools

Full param control for Claude, GPT-4o, Gemini. Temperature, top_p, streaming, extended thinking. BYOK.

Multi-Model Compare

One prompt → Claude + GPT-4o + Gemini at once. Speed, cost, output. No official console does this.

Paste any text → token count across all models. Context window fill bar. No API key needed.

Cost Calculator

Monthly cost across all providers. Full pricing table. Find cheapest model for your use case.

[02]Prompt Tools

3 tools

Save, version, diff prompts across providers. Rollback any version. One click to test in workbench.

Caching Calculator

See exactly how much prompt caching saves. "Cache this prefix → save $3,400/month." Free.

Streaming Debugger

Live tokens/sec, TTFT measurement, stall detection. Visibility nobody else offers.

[03]Reference

2 tools

Provider Status

Real-time uptime for Anthropic, OpenAI, Gemini, Groq. One page instead of 4 status sites.

Pricing, rate limits, context windows, benchmarks. All providers, one place.

[04]AI Ecosystem

curated links

→ Model Providers & Hubs

→ Dev Tools & Agents

→ Orchestration & SDKs

→ Local AI & Inference

→ Vector Databases

→ Observability

→ Consumer AI

[05]Documentation Hub

Manual & Help

Core Tools

Multi-Model Compare

Cost Calculator

Prompt Tools

Prompt Optimizer

Caching Calculator

Streaming Debugger

Reference

Provider Status