AI Token Price Intelligence & Smart Routing Platform for Agents
OhMyToken has 5 well-named tools with clear descriptions and structured schemas using Zod. Naming is excellent (verb_noun convention: get_token_price, compare_models, recommend_model, calculate_cost, get_price_trends). Descriptions are present and actionable (avg ~100 chars, meeting the 10-1024 baseline). All parameters have type definitions and descriptions. However, output schemas are NOT documented, responses are constructed as plain text strings without explicit return type schemas. Error handling exists but lacks actionable recovery guidance. No pagination support despite tools that could return large result sets. Security is adequate (read-only operations, no secrets exposed). Composition is sound, each tool has a single responsibility. Overall: solid community-grade implementation with room for structure maturity.
Calculate the estimated cost for a specific workload across one or all models.
Compare token pricing and capabilities between two or more AI models side by side.
Get historical token price trends showing how AI model pricing has changed over time.
Get current token pricing for a specific AI model or provider. Returns input/output price per million tokens in USD.
Get a smart model recommendation based on your task type, budget, and preferences.
Output schemas not documented. Tools return text-only responses constructed as markdown strings with no structured schema definition. LLMs cannot plan downstream operations or extract typed fields.
No pagination support. Tools that could return all models (get_token_price without filters, calculate_cost without model param) return unbounded result sets as single text blocks. Violates pattern:paginated-result.
Error messages lack recovery guidance. When model lookup fails, the error states 'No models found' but does not suggest next steps (e.g. 'try search without filters' or 'see available models'). Minimal guidance for LLM recovery.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-22 | D | 58 | 2026-07-28+ | v2 |
| 2026-03-09 | D | 57 | - | v1 |
No tool annotations. Tools are all read-only (readOnlyHint=true) but this is not declared in the tool definitions. Missing destructiveHint and idempotentHint where applicable. Spec requires tool annotations for LLM safety reasoning.
Result limits not enforced or documented. If the data model grows to hundreds of models, calculate_cost with no model filter could return massive unstructured output. No mention of limits in descriptions.