From 09e106c48a597c51e9ce767e90093397cc7d9b04 Mon Sep 17 00:00:00 2001 From: fpagny Date: Fri, 24 Apr 2026 17:44:58 +0200 Subject: [PATCH 1/4] feat(genapi): add gemma4 to supported models --- pages/generative-apis/reference-content/supported-models.mdx | 1 + 1 file changed, 1 insertion(+) diff --git a/pages/generative-apis/reference-content/supported-models.mdx b/pages/generative-apis/reference-content/supported-models.mdx index 05f029d4ca..79af84895d 100644 --- a/pages/generative-apis/reference-content/supported-models.mdx +++ b/pages/generative-apis/reference-content/supported-models.mdx @@ -17,6 +17,7 @@ Our API supports the most popular models for [Chat](/generative-apis/how-to/quer |-----------------|-----------------|-----------------|-----------------|-----------------|-----------------| | Qwen | `qwen3.5-397b-a17b` | 250k | 16k | [Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0) | [HF](https://huggingface.co/Qwen/Qwen3.5-397B-A17B) | | Mistral | `mistral-small-3.2-24b-instruct-2506` | 128k | 32k | [Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0) | [HF](https://huggingface.co/mistralai/Mistral-Small-3.2-24B-Instruct-2506) | +| Google (Preview) | `gemma-4-26b-a4b-it` | 256k | 32k | [Apache-2.0](https://www.apache.org/licenses/LICENSE-2.0) | [HF](https://huggingface.co/google/gemma-4-26B-A4B-it) | | Google (Preview) | `gemma-3-27b-it` | 40k | 8k | [Gemma](https://ai.google.dev/gemma/terms) | [HF](https://huggingface.co/google/gemma-3-27b-it) | | H | `holo2-30b-a3b` | 22k | 32k | [CC-BY-NC-4.0](https://spdx.org/licenses/CC-BY-NC-4.0) | [HF](https://huggingface.co/Hcompany/Holo2-30B-A3B) | From bc06c71556bdf1859aaf4c257c5b71ef7f0af10d Mon Sep 17 00:00:00 2001 From: fpagny Date: Fri, 24 Apr 2026 18:03:13 +0200 Subject: [PATCH 2/4] feat(inference): add gemma4 model details Added Gemma-4-26b-a4b-it model details to the catalog. --- .../reference-content/model-catalog.mdx | 16 ++++++++++++++++ 1 file changed, 16 insertions(+) diff --git a/pages/managed-inference/reference-content/model-catalog.mdx b/pages/managed-inference/reference-content/model-catalog.mdx index 238f7e6801..aa543bd398 100644 --- a/pages/managed-inference/reference-content/model-catalog.mdx +++ b/pages/managed-inference/reference-content/model-catalog.mdx @@ -20,6 +20,7 @@ A quick overview of available models in Scaleway's catalog and their core attrib | [`whisper-large-v3`](#whisper-large-v3) | OpenAI | - | Audio transcription | L4, L40S, H100, H100-SXM-2 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`qwen3.5-397b-a17b`](#qwen35-397b-a17b) | Qwen | 250k | Text, Code, Vision | H100-SXM-8 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`qwen3-235b-a22b-instruct-2507`](#qwen3-235b-a22b-instruct-2507) | Qwen | 250k | Text | H100-SXM-2 (40k), H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | +| [`gemma-4-26b-a4b-it`](#gemma-4-26b-a4b-it) | Google | 250k | Text, Vision | H100-SXM-2, H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`gemma-3-27b-it`](#gemma-3-27b-it) | Google | 40k | Text, Vision | H100, H100-2 | [Gemma](https://ai.google.dev/gemma/terms) | | [`llama-3.3-70b-instruct`](#llama-33-70b-instruct) | Meta | 128k | Text | H100 (15k), H100-2 | [Llama 3.3 Community](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) | | [`llama-3.1-70b-instruct`](#llama-31-70b-instruct) | Meta | 128k | Text | H100 (15k), H100-2 | [Llama 3.1 Community](https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct/blob/main/LICENSE) | @@ -58,6 +59,7 @@ A quick overview of available models in Scaleway's catalog and their core attrib | `whisper-large-v3` | - | - | English, French, German, Chinese, Japanese, Korean and 81 additional languages | | `qwen3.5-397b-a17b` | Yes | Yes | English, French, German, Chinese, Japanese, Korean and 113 additional languages and dialects | | `qwen3-235b-a22b-instruct-2507` | Yes | Yes | English, French, German, Chinese, Japanese, Korean and 113 additional languages and dialects | +| `gemma-4-26b-a4b-it` | Yes | Yes | English, Chinese, Japanese, Korean and 136 additional languages | | `gemma-3-27b-it` | Yes | Partial | English, Chinese, Japanese, Korean and 31 additional languages | | `llama-3.3-70b-instruct` | Yes | Yes | English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai | | `llama-3.1-70b-instruct` | Yes | Yes | English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai | @@ -98,6 +100,20 @@ A quick overview of available models in Scaleway's catalog and their core attrib Vision models can understand and analyze images, not generate them. You will use it through the /v1/chat/completions endpoint. +### Gemma-4-26b-a4b-it +Gemma-4-26b-a4b-it is a frontier small model as of April 2026 to perform agentic and reasoning tasks on many languages. + +#### Model names +``` +google/gemma-4-26b-a4b-it +``` +| Attribute | Value | +|-----------|-------| +| Supports parallel tool calling | Yes | +| Supported image formats | PNG, JPEG, WEBP, and non-animated GIFs | +| Maximum image resolution (pixels) | 896x896 | +| Token dimension (pixels)| 64x64 | + ### Gemma-3-27b-it Gemma-3-27b-it is a model developed by Google to perform text processing and image analysis on many languages. The model was not trained specifically to output function / tool call tokens. Hence function calling is currently supported, but reliability remains limited. From a892a35ea2bc5712607090d7d2d376a425772182 Mon Sep 17 00:00:00 2001 From: fpagny Date: Fri, 24 Apr 2026 18:09:09 +0200 Subject: [PATCH 3/4] feat(inference): add reasoning efforts parameters Added supported reasoning efforts to model attributes. --- pages/managed-inference/reference-content/model-catalog.mdx | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) diff --git a/pages/managed-inference/reference-content/model-catalog.mdx b/pages/managed-inference/reference-content/model-catalog.mdx index aa543bd398..eb21d797ab 100644 --- a/pages/managed-inference/reference-content/model-catalog.mdx +++ b/pages/managed-inference/reference-content/model-catalog.mdx @@ -105,11 +105,12 @@ Gemma-4-26b-a4b-it is a frontier small model as of April 2026 to perform agentic #### Model names ``` -google/gemma-4-26b-a4b-it +google/gemma-4-26b-a4b-it:bf16 ``` | Attribute | Value | |-----------|-------| | Supports parallel tool calling | Yes | +| Supported reasoning efforts | `none`, `low`, `medium`, `high` | | Supported image formats | PNG, JPEG, WEBP, and non-animated GIFs | | Maximum image resolution (pixels) | 896x896 | | Token dimension (pixels)| 64x64 | @@ -191,6 +192,7 @@ This model was released as a frontier reasoning model on 16th February 2026. | Attribute | Value | |-----------|-------| | Supports parallel tool calling | Yes | +| Supported reasoning efforts | `none`, `low`, `medium`, `high` | | Supported images formats | PNG, JPEG, WEBP, and non-animated GIFs | | Supported video formats | MP4, MPEG, MOV, OGG and WEBM | | Maximum image resolution (pixels) | 4096x4096 | @@ -309,6 +311,7 @@ Currently, this model should be used through Responses API, as Chat Completion d | Attribute | Value | |-----------|-------| | Supports parallel tool calling | Yes | +| Supported reasoning efforts | `low`, `medium`, `high` | #### Model name ``` From b17b427e2870d7966255d31dc711dbbc21520f94 Mon Sep 17 00:00:00 2001 From: fpagny Date: Mon, 27 Apr 2026 13:59:12 +0200 Subject: [PATCH 4/4] feat(genapi): add gemma 4 26b a4b Updated Gemma model entries to reflect new version and attributes. --- .../reference-content/model-catalog.mdx | 20 ++++++++++++++++++- 1 file changed, 19 insertions(+), 1 deletion(-) diff --git a/pages/managed-inference/reference-content/model-catalog.mdx b/pages/managed-inference/reference-content/model-catalog.mdx index eb21d797ab..fe289ca595 100644 --- a/pages/managed-inference/reference-content/model-catalog.mdx +++ b/pages/managed-inference/reference-content/model-catalog.mdx @@ -20,7 +20,8 @@ A quick overview of available models in Scaleway's catalog and their core attrib | [`whisper-large-v3`](#whisper-large-v3) | OpenAI | - | Audio transcription | L4, L40S, H100, H100-SXM-2 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`qwen3.5-397b-a17b`](#qwen35-397b-a17b) | Qwen | 250k | Text, Code, Vision | H100-SXM-8 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`qwen3-235b-a22b-instruct-2507`](#qwen3-235b-a22b-instruct-2507) | Qwen | 250k | Text | H100-SXM-2 (40k), H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | -| [`gemma-4-26b-a4b-it`](#gemma-4-26b-a4b-it) | Google | 250k | Text, Vision | H100-SXM-2, H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | +| [`gemma-4-31b-it`](#gemma-4-31b-it) | Google | 250k | Text, Vision | H100 (60k), H100-SXM-2, H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | +| [`gemma-4-26b-a4b-it`](#gemma-4-26b-a4b-it) | Google | 250k | Text, Vision | H100, H100-SXM-2, H100-SXM-4 | [Apache 2.0](https://www.apache.org/licenses/LICENSE-2.0) | | [`gemma-3-27b-it`](#gemma-3-27b-it) | Google | 40k | Text, Vision | H100, H100-2 | [Gemma](https://ai.google.dev/gemma/terms) | | [`llama-3.3-70b-instruct`](#llama-33-70b-instruct) | Meta | 128k | Text | H100 (15k), H100-2 | [Llama 3.3 Community](https://huggingface.co/meta-llama/Llama-3.3-70B-Instruct) | | [`llama-3.1-70b-instruct`](#llama-31-70b-instruct) | Meta | 128k | Text | H100 (15k), H100-2 | [Llama 3.1 Community](https://huggingface.co/meta-llama/Llama-3.1-70B-Instruct/blob/main/LICENSE) | @@ -59,6 +60,7 @@ A quick overview of available models in Scaleway's catalog and their core attrib | `whisper-large-v3` | - | - | English, French, German, Chinese, Japanese, Korean and 81 additional languages | | `qwen3.5-397b-a17b` | Yes | Yes | English, French, German, Chinese, Japanese, Korean and 113 additional languages and dialects | | `qwen3-235b-a22b-instruct-2507` | Yes | Yes | English, French, German, Chinese, Japanese, Korean and 113 additional languages and dialects | +| `gemma-4-31b-it` | Yes | Yes | English, Chinese, Japanese, Korean and 136 additional languages | | `gemma-4-26b-a4b-it` | Yes | Yes | English, Chinese, Japanese, Korean and 136 additional languages | | `gemma-3-27b-it` | Yes | Partial | English, Chinese, Japanese, Korean and 31 additional languages | | `llama-3.3-70b-instruct` | Yes | Yes | English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai | @@ -100,8 +102,24 @@ A quick overview of available models in Scaleway's catalog and their core attrib Vision models can understand and analyze images, not generate them. You will use it through the /v1/chat/completions endpoint. +### Gemma-4-31b-it +Gemma-4-31b-it is a frontier small model as of April 2026 to perform agentic and reasoning tasks on many languages. + +#### Model names +``` +google/gemma-4-31b-it:bf16 +``` +| Attribute | Value | +|-----------|-------| +| Supports parallel tool calling | Yes | +| Supported reasoning efforts | `none`, `low`, `medium`, `high` | +| Supported image formats | PNG, JPEG, WEBP, and non-animated GIFs | +| Maximum image resolution (pixels) | 896x896 | +| Token dimension (pixels)| 64x64 | + ### Gemma-4-26b-a4b-it Gemma-4-26b-a4b-it is a frontier small model as of April 2026 to perform agentic and reasoning tasks on many languages. +This model has a Mixture-of-Expert (MoE) architecture, providing significant throughput and fitting on a single H100 GPU while supporting its maximum context size. #### Model names ```