Commit Graph

51 Commits

Author SHA1 Message Date
Premshay
8460a796e9 Merge branch 'main' into main 2024-12-15 11:15:43 +02:00
Matt Rubens
6ee118e0a2 Merge pull request #86 from RooVetGit/fix-openai-compatible-streaming
Add 'Include stream options' checkbox for OpenAI-compatible providers
2024-12-13 01:41:15 -05:00
Matt Rubens
be3d8a6166 Incorporate MCP changes (#93)
Co-authored-by: Saoud Rizwan <7799382+saoudrizwan@users.noreply.github.com>
2024-12-12 23:16:39 -05:00
Matt Rubens
2cfd76c305 Add 'Include stream options' checkbox for OpenAI-compatible providers 2024-12-12 09:20:58 -05:00
Matt Rubens
156cab8cbc Fix: Add checkbox for stream_options in OpenAI-compatible providers 2024-12-12 08:58:38 -05:00
Cline
51a57d5bbf fix(bedrock): improve stream handling and type safety
- Fix TypeScript error in ConverseStreamCommand payload
- Add proper JSON parsing for test stream events
- Improve error handling with proper Error objects
- Add test-specific model info with required fields
- Fix cross-region inference and prompt cache config
2024-12-11 22:47:59 +02:00
Cline
140318cecd feat(api): unify Bedrock provider using Runtime API
Problem:
The current Bedrock implementation uses the Bedrock SDK, which requires separate handling for different model types and doesn't provide a unified streaming interface.

Solution:
Integrate the Bedrock Runtime API to provide a single, unified interface for all Bedrock models (Claude and Nova) using the ConverseStream API. This eliminates the need for separate handlers while maintaining all existing functionality.

Key Changes:
- Refactored AwsBedrockHandler to use @aws-sdk/client-bedrock-runtime
- Enhanced bedrock-converse-format.ts to handle all content types and properly transform between Anthropic and Bedrock formats
- Maintained cross-region inference support with proper region prefixing
- Added support for prompt caching configuration
- Improved AWS credentials handling to better support default providers
- Added proper error handling and token tracking for all response types

Dependencies:
- Added @aws-sdk/client-bedrock-runtime for unified API access
- Removed @anthropic-ai/bedrock-sdk dependency

Testing:
- Verified message format conversion for all content types
- Tested cross-region inference functionality
- Validated streaming responses for both Claude and Nova models

This change simplifies the codebase by providing a single, consistent interface for all Bedrock models while maintaining full compatibility with existing features.
2024-12-11 22:47:56 +02:00
Cline
dffc040e7c feat(bedrock): Add Meta Llama 3, 3.1, and 3.2 models with detailed pricing and context windows 2024-12-11 22:31:32 +02:00
Matt Rubens
0e0c848a96 Fix Gemini 2.0 context window 2024-12-11 14:02:54 -05:00
Matt Rubens
65bbefed55 Add Gemini 2.0 2024-12-11 13:36:46 -05:00
Domenico Bochicchio | Asclepyus Srl
39b51fa78f Added support for gemini-exp-1206 (#54) 2024-12-09 09:37:38 -05:00
John Stearns
423e2af520 Continuing work on support for OpenRouter compression (#43) 2024-12-07 09:38:13 -08:00
Saoud Rizwan
39bc35eec1 Add LM Studio provider 2024-11-12 22:02:42 -05:00
Saoud Rizwan
ad29ff2a03 Add AWS cross-region inference toggle 2024-11-07 13:51:13 -05:00
Sam
2eb11aadc7 Fix: Add missing sonnet model to bedrock (#633) 2024-11-07 12:57:03 -05:00
Saoud Rizwan
48cdfa9874 Add Haiku 3.5 to Anthropic, Bedrock, Vertex 2024-11-04 14:08:25 -05:00
Saoud Rizwan
17f34e6225 Add 'supports computer use' prop; sync latest openrouter info on launch 2024-10-23 15:43:34 -04:00
Saoud Rizwan
e5266ddda1 Add fallback claude 3.5 sonnet v1 models for bedrock and vertex 2024-10-22 18:24:34 -04:00
Saoud Rizwan
165272ab47 Update Anthropic model IDs 2024-10-22 16:21:17 -04:00
Saoud Rizwan
7dd9abda46 Cleanup 2024-10-09 01:49:59 -04:00
Saoud Rizwan
f6a14fdfb9 Add markdown rendering for model description; fixes 2024-10-09 01:49:57 -04:00
Saoud Rizwan
7cb0c524e5 Add OpenRouter custom model scheme 2024-10-09 01:49:57 -04:00
Saoud Rizwan
5069e1a11b Use new gemini models 2024-09-25 00:28:18 -04:00
Saoud Rizwan
d81002decc Refactor Azure OpenAI default version 2024-09-24 12:45:24 -04:00
Saoud Rizwan
abb4a0143e Add Azure API version option 2024-09-20 10:56:26 -04:00
Saoud Rizwan
adceb76775 Add o1 support 2024-09-13 16:37:04 -04:00
Saoud Rizwan
e2d62ab109 Add experimental gemini models 2024-09-13 04:41:58 -04:00
Saoud Rizwan
dc617a92a9 Add prompt caching to openrouter; remove o1 since it doesn't support tool use 2024-09-13 00:49:04 -04:00
Saoud Rizwan
4b44e8f921 Add OpenAI provider 2024-09-12 15:01:28 -04:00
Saoud Rizwan
3b004aed37 Fix gemini message conversion 2024-09-12 11:49:32 -04:00
Saoud Rizwan
fbb7620fa1 Add gemini support 2024-09-12 08:11:33 -04:00
Saoud Rizwan
ca4ad1d0b3 Add anthropicBaseUrl option 2024-09-07 07:31:29 -04:00
Saoud Rizwan
9eaf753000 Add ollama models polling 2024-09-06 04:51:17 -04:00
Saoud Rizwan
286e569e09 Add ollama provider option 2024-09-03 23:03:30 -04:00
Saoud Rizwan
c209198b23 Add openai compatible provider 2024-09-03 17:08:29 -04:00
Saoud Rizwan
0badfa2706 Add support for aws credentials file or environment variables, and session token 2024-09-03 13:58:21 -04:00
Saoud Rizwan
af46531e52 Minor fixes 2024-08-28 06:13:46 -04:00
Saoud Rizwan
13af5992af Add GCP Vertex AI provider 2024-08-28 05:35:49 -04:00
Saoud Rizwan
843ef29a07 Remove Kodu provider 2024-08-27 21:38:01 -04:00
Saoud Rizwan
650283a11f Add sliding window context management 2024-08-26 01:00:00 -04:00
Saoud Rizwan
5dcf836d81 Enable prompt caching for Opus 2024-08-26 01:00:00 -04:00
Saoud Rizwan
df4e8e7afc Add Kodu provider 2024-08-26 01:00:00 -04:00
Saoud Rizwan
f6fd76823b Add Maestro login button 2024-08-26 01:00:00 -04:00
Saoud Rizwan
6989779dd6 Update announcement/settings with details about prompt caching 2024-08-15 04:21:28 -04:00
Saoud Rizwan
ec2bfa352a Add prompt cache 2024-08-15 03:34:54 -04:00
Saoud Rizwan
e00b63ef03 Use OpenRouter's self-moderated claude models 2024-08-12 23:35:38 -04:00
Saoud Rizwan
f93e7946aa Remove unusable models; default to OpenRouter for new users; update Announcement 2024-08-11 15:37:09 -04:00
Saoud Rizwan
6454163078 Remove unusable models; minor fixes 2024-08-11 03:24:10 -04:00
Saoud Rizwan
3c082e753b Add new models 2024-08-11 02:32:54 -04:00
Saoud Rizwan
f54774b943 Add option to choose different models 2024-08-11 00:28:22 -04:00