Behind the scenes, with made-up data. This is a read only copy of the admin for a fictional candidate. Reports, traces, spend and chat were produced by a scripted stand-in; the evaluation timeline shows real results of real models on fictional cases. Nothing here can be changed. Back to the real site
Reverse Apply

Models

Which models may run, which one does which job, and what each has cost and scored so far.

Cost against quality

Cost against quality, by model1 model. Across: mean cost of one finished check in US dollars. Up: share of synthetic eval cases passed. 0 left out because a cost or an eval result is missing. The same numbers are in the catalog table.050100$0$0.0000$0.0001Mean cost of one finished checkSynthetic eval cases passed, %claude-haiku-5.5: $0.0001 a report, 89% of eval cases passedclaude-haiku-5.5

Quality here is the share of cases passed on the synthetic datasets only, where the right answer follows from how each case is written. Other kinds are not mixed in. Only models with a cost on record and a finished run of that kind are drawn. Up and to the left is better.

Roles

Reports, chat and letters for people who paste a posting. It must have a price, so a local model will not do. Only models up to $1.0 in / $5.0 out per million tokens are listed (the ceiling is under Settings). None turns checks for visitors off.

None means the model that answers visitors.

Reads a verdict and its cited CV line and says whether the line supports it.

Catalog

11 of 11 switched on

Fetch the current list and prices from

Bedrock models are hidden until this server has AWS credentials.

ModelStatePrice per million tokens, in / outReasoning effortReportsEval cases passed, by kindTime, median
Baseline (keyword overlap, no model)baseline On
Key set, can run		 No model, no cost not used 335 ms
Claude haiku (local `claude` command, plan usage)cli:haiku On
Not ready: `claude` command not found		 Development only, not billed not used 0
Claude sonnet (local `claude` command, plan usage)cli:sonnet On
Not ready: `claude` command not found		 Development only, not billed not used 0
Claude Haiku 5.5 (OpenRouter)openrouter:anthropic/claude-haiku-5.5 On
Not ready: OPENROUTER_API_KEY is not set		 $0.10 / $0.50
4synthetic89%390 ms
DeepSeek V4.1 Flash (OpenRouter)openrouter:deepseek/deepseek-v4.1-flash-1017 On
Not ready: OPENROUTER_API_KEY is not set		 $0.30 / $1.20
0synthetic100%
Meta Muse Ember (OpenRouter)openrouter:meta/muse-ember On
Not ready: OPENROUTER_API_KEY is not set		 $0.10 / $0.50
0
NVIDIA Nemotron 3 Super 120B (OpenRouter)openrouter:nvidia/nemotron-3-super-120b-a12b On
Not ready: OPENROUTER_API_KEY is not set		 $0.07 / $0.40
0
NVIDIA Nemotron 3.5 Super VL (OpenRouter)openrouter:nvidia/nemotron-3.5-super-vl On
Not ready: OPENROUTER_API_KEY is not set		 $0.15 / $0.47
0
OpenAI GPT-6.1 Sol (OpenRouter)openrouter:openai/gpt-6.1-sol On
Not ready: OPENROUTER_API_KEY is not set		 $2.00 / $10.00
0
OpenAI GPT-6.1 Terra (OpenRouter)openrouter:openai/gpt-6.1-terra On
Not ready: OPENROUTER_API_KEY is not set		 $1.00 / $5.00
0synthetic89%
Xiaomi MiMo V2.6 Pro (OpenRouter)openrouter:xiaomi/mimo-v2.6-pro On
Not ready: OPENROUTER_API_KEY is not set		 $0.43 / $0.87
0