add helicone models + helicone model generation script #325
Conversation
|
can we get a review on this? this PR is a prereq needed to add helicone support the opencode integration |
|
yes |
There was a problem hiding this comment.
can you mkae this follow the pattern of other logo.svg files?
There was a problem hiding this comment.
mostly just using: currentColor instead of actual colors
There was a problem hiding this comment.
heyo! saw that the logos are generally monochromatic, so I just left the helicone logo's outline and used the currentColor for the stroke color
There was a problem hiding this comment.
Pull request overview
This PR adds the Helicone AI gateway provider to the repository, including 85 model configurations and an automated generation script that fetches model data from Helicone's public API endpoint.
Key Changes:
- Introduces a TypeScript generation script that dynamically creates model TOML files from Helicone's model registry API
- Adds provider configuration and logo for Helicone
- Generates 85 model TOML files covering various AI models (OpenAI, Anthropic, Google, xAI, Meta, DeepSeek, etc.)
Reviewed changes
Copilot reviewed 95 out of 96 changed files in this pull request and generated 56 comments.
Show a summary per file
| File | Description |
|---|---|
| providers/helicone/provider.toml | Provider configuration defining Helicone gateway API endpoint and environment variables |
| providers/helicone/README.md | Documentation explaining the generation script, commands, and data format conventions |
| providers/helicone/logo.svg | Helicone brand logo in SVG format |
| providers/helicone/models/*.toml | 85 model configuration files with pricing, capabilities, and limits (auto-generated) |
| packages/core/script/generate-helicone.ts | TypeScript script to fetch and transform Helicone's model registry into TOML format |
| package.json | Adds helicone:generate npm script to run the model generation |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
|
|
||
| [cost] | ||
| input = 0.3 | ||
| output = 2.9000000000000004 |
There was a problem hiding this comment.
The output cost has excessive floating-point precision (2.9000000000000004) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 2.9.
| [cost] | ||
| input = 1.25 | ||
| output = 10 | ||
| cache_read = 0.12500000000000003 |
There was a problem hiding this comment.
The cache_read cost has excessive floating-point precision (0.12500000000000003) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.125.
| [cost] | ||
| input = 5 | ||
| output = 25 | ||
| cache_read = 0.5000000000000001 |
There was a problem hiding this comment.
The cache_read cost has excessive floating-point precision (0.5000000000000001) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.5.
| cache_read = 0.5000000000000001 | |
| cache_read = 0.5 |
| [cost] | ||
| input = 3 | ||
| output = 15 | ||
| cache_read = 0.30000000000000004 |
There was a problem hiding this comment.
The cache_read cost has excessive floating-point precision (0.30000000000000004) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.3.
| cache_read = 0.30000000000000004 | |
| cache_read = 0.3 |
| open_weights = false | ||
|
|
||
| [cost] | ||
| input = 0.19999999999999998 |
There was a problem hiding this comment.
The input cost has excessive floating-point precision (0.19999999999999998) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.2.
| open_weights = false | ||
|
|
||
| [cost] | ||
| input = 0.19999999999999998 |
There was a problem hiding this comment.
The input cost has excessive floating-point precision (0.19999999999999998) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.2.
| [cost] | ||
| input = 1.25 | ||
| output = 10 | ||
| cache_read = 0.12500000000000003 |
There was a problem hiding this comment.
The cache_read cost has excessive floating-point precision (0.12500000000000003) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.125.
|
|
||
| [cost] | ||
| input = 0.049999999999999996 | ||
| output = 0.09999999999999999 |
There was a problem hiding this comment.
The output cost has excessive floating-point precision (0.09999999999999999) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.1.
| [cost] | ||
| input = 1 | ||
| output = 5 | ||
| cache_read = 0.09999999999999999 |
There was a problem hiding this comment.
The input cost has excessive floating-point precision (0.09999999999999999) which results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.1.
| cache_read = 0.09999999999999999 | |
| cache_read = 0.1 |
| open_weights = false | ||
|
|
||
| [cost] | ||
| input = 0.09999999999999999 |
There was a problem hiding this comment.
The input cost has excessive floating-point precision (0.09999999999999999) which likely results from JavaScript's floating-point arithmetic. The value should be rounded to a reasonable precision, e.g., 0.1.
2 participants
Hi! Wanted to add the helicone gateway to the repo.
Their models can be retrieved through an endpoint, so I thought for maintainability it would be good to have a script that updates the database from the endpoint, for the most accurate information.
If this is not preferred please let me know and I will have the script removed!