Tested 2026-10-15
Scoring inbound leads from a contact form: inclusionAI: Ling 3.0 Flash won, 3 of 7 models failed a must-have; OpenRouter's own picker passed every must-have but cost 8× the winner (4th of 4)
Read contact-form leads and say whether each fits the agency's criteria, with a one-line reason — for a 5-person web design agency.
What this project actually needed
- takes written text as input — must have
- produces written text — must have
- applies the stated criteria and nothing else — must have: every verdict follows the stated criteria
- cost at your volume — decides the order
The winner
inclusionAI: Ling 3.0 Flash passed every must-have and ranked first by how this job said to choose.
OpenRouter's automatic picker sent this job to google/gemini-2.5-flash, z-ai/glm-5.3-flash and passed every must-have but cost 8× the winner (4th of 4). It picks one message at a time and never sees this job's must-haves.
Who failed, in their own words
Mistral: Mistral Nemo
- applies the stated criteria and nothing else:
NOT FIT - Budget is below $5,000 requirement.
- facts match the task:
Budget is below $5,000 requirement
Meta: Llama 3.1 8B Instruct
- applies the stated criteria and nothing else:
However, I don't see a specific company or lead you'd like me to assess as either FIT or NOT FIT based on the given budget and location criteria.
- facts match the task:
$5000 budget is above the minimum required
inclusionAI: Ling 3.0 Flash VL
- gives an answer:
no answer came back after three tries (HTTP 429: b'{"error":{"message":"Provider returned error","c)
Every model on every requirement
| Model | cost at your volume | produces written text | takes written text as input | applies the stated criteria and nothing else |
|---|---|---|---|---|
| Auto Router | $0.0017 a day (measured on your tasks) | not applicable — the picker chooses a model per request | not applicable — the picker chooses a model per request | passed 3 of 3 tasks |
| OpenAI: gpt-oss-20b | $0.0003 a day (measured on your tasks) | yes (estimate — not observed) | yes (estimate — not observed) | passed 3 of 3 tasks |
| Mistral: Mistral Nemo | $0.0000 a day (measured on your tasks) | yes (estimate — not observed) | yes (estimate — not observed) | passed 1 of 3 tasks |
| Meta: Llama 3.1 8B Instruct | $0.0001 a day (measured on your tasks) | yes (estimate — not observed) | yes (estimate — not observed) | passed 1 of 3 tasks |
| inclusionAI: Ling 3.0 Flash | $0.0002 a day (measured on your tasks) | yes (estimate — not observed) | yes (estimate — not observed) | passed 3 of 3 tasks |
| inclusionAI: Ling 3.1 Flash | $0.0010 a day (measured on your tasks) | yes (estimate — not observed) | yes (estimate — not observed) | passed 3 of 3 tasks |
| inclusionAI: Ling 3.0 Flash VL | not measured | yes (estimate — not observed) | yes (estimate — not observed) | not judged — no graded answers |
7 models tested on lead qualification; cost of this test $0.0367.
Your app does something like this? XF Model Watch keeps it on the best model. XF Model Watch re-tests your own job as new models arrive and emails you only when one passes every check. Watch this job