Tested 2026-10-27
Translating product listings into Spanish: Z.ai: GLM 5.3 Flash won, 3 of 8 models failed a must-have
Translate e-commerce product listings from English to Spanish, keeping sizes, prices and brand names exact — for an online store expanding to Mexico.
What this project actually needed
- takes written text as input — must have
- produces written text — must have
- keeps every number, size, price and brand name exactly — must have: nothing changed or dropped
- reads naturally in Spanish — decides the order
- cost at your volume — decides the order
The winner
Z.ai: GLM 5.3 Flash passed every must-have and ranked first by how this job said to choose.
OpenRouter's automatic picker sent this job to openai/gpt-6-luna and was not ranked. It picks one message at a time and never sees this job's must-haves.
Who failed, in their own words
Auto Router
- keeps every number, size, price and brand name exactly:
24 $
ScaleDown: Summarize
- keeps every number, size, price and brand name exactly:
24 dólares
Anthropic: Claude Opus 4.7
- keeps every number, size, price and brand name exactly:
6-12
Every model on every requirement
| Model | cost at your volume | produces written text | reads naturally in Spanish | takes written text as input | keeps every number, size, price and brand name exactly |
|---|---|---|---|---|---|
| Auto Router | $0.0137 a month (measured on your tasks) | not applicable — the picker chooses a model per request | scored 1.00 of 1 across 6 tasks (graded) | not applicable — the picker chooses a model per request | passed 4 of 6 tasks |
| Z.ai: GLM 5.3 Flash | $0.1133 a month (measured on your tasks) | yes (estimate — not observed) | scored 1.00 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 3 of 3 tasks |
| Meta: Muse Spark 1.1 | $1.2635 a month (measured on your tasks) | yes (estimate — not observed) | scored 1.00 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 3 of 3 tasks |
| ScaleDown: Summarize | $0.0030 a month (measured on your tasks) | yes (estimate — not observed) | scored 1.00 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 2 of 3 tasks |
| Anthropic: Claude Opus 4.6 | $0.6200 a month (measured on your tasks) | yes (estimate — not observed) | scored 1.00 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 3 of 3 tasks |
| Anthropic: Claude Opus 4.7 | $2.1283 a month (measured on your tasks) | yes (estimate — not observed) | scored 0.97 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 2 of 3 tasks |
| Google: Gemini 3.1 Pro Preview | $4.3733 a month (measured on your tasks) | yes (estimate — not observed) | scored 1.00 of 1 across 3 tasks (graded) | yes (estimate — not observed) | passed 3 of 3 tasks |
8 models tested on listing translation spanish; cost of this test $0.0822.
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