github-starred/ollama
2
0
Fork 0
mirror of https://github.com/ollama/ollama.git synced 2026-05-06 16:11:34 -05:00
Code Issues 14.5k Packages Projects Releases 120 Wiki Activity

[PR #15243] api: add audios field for audio input in multimodal models #25630

New Issue
Open
opened 2026-04-19 18:19:16 -05:00 by GiteaMirror · 0 comments
GiteaMirror commented 2026-04-19 18:19:16 -05:00
Owner
Copy Link

📋 Pull Request Information

Original PR: https://github.com/ollama/ollama/pull/15243
Author: @harsha-gouru
Created: 4/2/2026
Status: 🔄 Open

Base: main ← Head: feat/audio-api-field

📝 Commits (1)

  • af45c40 api: add audios field for audio input in multimodal models

📊 Changes

15 files changed (+495 additions, -27 deletions)

View changed files

📝 api/types.go (+6 -1)
📝 cmd/cmd.go (+10 -1)
📝 cmd/cmd_test.go (+9 -0)
📝 cmd/interactive.go (+8 -5)
📝 cmd/interactive_test.go (+5 -3)
📝 integration/audio_test.go (+2 -2)
📝 model/renderers/gemma4.go (+4 -3)
📝 openai/openai.go (+5 -2)
📝 openai/openai_test.go (+201 -0)
📝 server/prompt.go (+9 -3)
📝 server/prompt_test.go (+26 -0)
📝 server/routes.go (+12 -6)
📝 server/routes_debug_test.go (+35 -0)
📝 server/routes_generate_test.go (+157 -0)
📝 template/template.go (+6 -1)

📄 Description

Add a dedicated Audios field to Message and GenerateRequest so that API consumers can send audio data without stuffing it into the Images field. Audio bytes are merged into the existing multimodal pipeline at the server layer since models already detect WAV format by magic bytes.

Changes:

  • Add Audios []ImageData to GenerateRequest and Message structs
  • Merge audios into the images pipeline in prompt.go and routes.go
  • Update Gemma4 renderer and template rendering to count audios
  • Route OpenAI input_audio content to the new Audios field
  • Validate input_audio format and reject non-WAV with a clear error
  • Fix audio-only generate being treated as a load-only request
  • Separate WAV files from images in CLI extractFileData
  • Add server, OpenAI, prompt, and debug-render tests for audio

Backward compatible: the images field continues to accept audio data.

Fixes #11798

🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.

## 📋 Pull Request Information **Original PR:** https://github.com/ollama/ollama/pull/15243 **Author:** [@harsha-gouru](https://github.com/harsha-gouru) **Created:** 4/2/2026 **Status:** 🔄 Open **Base:** `main` ← **Head:** `feat/audio-api-field` --- ### 📝 Commits (1) - [`af45c40`](https://github.com/ollama/ollama/commit/af45c407cece58d9e46e3c86236b857665165608) api: add audios field for audio input in multimodal models ### 📊 Changes **15 files changed** (+495 additions, -27 deletions) <details> <summary>View changed files</summary> 📝 `api/types.go` (+6 -1) 📝 `cmd/cmd.go` (+10 -1) 📝 `cmd/cmd_test.go` (+9 -0) 📝 `cmd/interactive.go` (+8 -5) 📝 `cmd/interactive_test.go` (+5 -3) 📝 `integration/audio_test.go` (+2 -2) 📝 `model/renderers/gemma4.go` (+4 -3) 📝 `openai/openai.go` (+5 -2) 📝 `openai/openai_test.go` (+201 -0) 📝 `server/prompt.go` (+9 -3) 📝 `server/prompt_test.go` (+26 -0) 📝 `server/routes.go` (+12 -6) 📝 `server/routes_debug_test.go` (+35 -0) 📝 `server/routes_generate_test.go` (+157 -0) 📝 `template/template.go` (+6 -1) </details> ### 📄 Description Add a dedicated Audios field to Message and GenerateRequest so that API consumers can send audio data without stuffing it into the Images field. Audio bytes are merged into the existing multimodal pipeline at the server layer since models already detect WAV format by magic bytes. Changes: - Add Audios []ImageData to GenerateRequest and Message structs - Merge audios into the images pipeline in prompt.go and routes.go - Update Gemma4 renderer and template rendering to count audios - Route OpenAI input_audio content to the new Audios field - Validate input_audio format and reject non-WAV with a clear error - Fix audio-only generate being treated as a load-only request - Separate WAV files from images in CLI extractFileData - Add server, OpenAI, prompt, and debug-render tests for audio Backward compatible: the images field continues to accept audio data. Fixes #11798 --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>
GiteaMirror added the pull-request label 2026-04-19 18:19:16 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
amd
api
app
bug
build
cli
cloud
compatibility
context-length
create
docker
documentation
embeddings
feature request
feedback wanted
good first issue
gpt-oss
gpu
harmony
help wanted
image
install
intel
js
launch
linux
macos
memory
mlx
model
needs more info
networking
nvidia
ollama.com
performance
pull-request
Mirrored from GitHub Pull Request
 python
question
registry
rendering
thinking
tools
top
vulkan
windows
wsl
No Label pull-request
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
GiteaMirror ninjasurge
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/ollama#25630
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?