From c7cf3cafe79ed79a6e91e13d97c722f6be7440bf Mon Sep 17 00:00:00 2001 From: Changelog bot Date: Wed, 25 Jun 2025 15:32:12 +0000 Subject: [PATCH 1/2] feat(changelog): add new entry --- ...-apis-changed-llama-33-70b-maximum-context-up.mdx | 12 ++++++++++++ 1 file changed, 12 insertions(+) create mode 100644 changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx diff --git a/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx b/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx new file mode 100644 index 0000000000..7d2ea96afa --- /dev/null +++ b/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx @@ -0,0 +1,12 @@ +--- +title: Llama 3.3 70B maximum context update +status: changed +date: 2025-06-25 +category: ai-data +product: generative-apis +--- + +Llama 3.3 70B maximum context is [now reduced to 100k tokens](https://www.scaleway.com/en/docs/generative-apis/reference-content/supported-models/) (from 130k tokens previously). +This update will improve average throughput and time to first token. +[Managed Inference](https://www.scaleway.com/en/docs/managed-inference/reference-content/model-catalog/) can still be used to support 130k tokens context length. + From 134f6b2ff15d41d62cb52bca21ffc883ef21e319 Mon Sep 17 00:00:00 2001 From: Benedikt Rollik Date: Thu, 26 Jun 2025 10:49:01 +0200 Subject: [PATCH 2/2] Update changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx Co-authored-by: Rowena Jones <36301604+RoRoJ@users.noreply.github.com> --- ...-generative-apis-changed-llama-33-70b-maximum-context-up.mdx | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx b/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx index 7d2ea96afa..0bbe3be2c9 100644 --- a/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx +++ b/changelog/june2025/2025-06-25-generative-apis-changed-llama-33-70b-maximum-context-up.mdx @@ -8,5 +8,5 @@ product: generative-apis Llama 3.3 70B maximum context is [now reduced to 100k tokens](https://www.scaleway.com/en/docs/generative-apis/reference-content/supported-models/) (from 130k tokens previously). This update will improve average throughput and time to first token. -[Managed Inference](https://www.scaleway.com/en/docs/managed-inference/reference-content/model-catalog/) can still be used to support 130k tokens context length. +[Managed Inference](https://www.scaleway.com/en/docs/managed-inference/reference-content/model-catalog/) can still be used to support context lengths of 130k tokens.