Releasesby GlyphMonitor your dedicated inference endpoints with the new Metrics API, providing Prometheus-compatible metrics...
Monitor your dedicated inference endpoints with the new Metrics API, providing Prometheus-compatible metrics for requests, tokens, latency percentiles, and endpoint health. Metrics are aggregated at the minute level and integrate directly with Prometheus. Learn more in our Metrics guide.
Also shipped Jan 22, 2026Cerebras in January 2026 →
Visualization suggestions updates
Visualization Suggestions , which were introduced in Grafana 8.3.0 , have been updated to provide more applicable suggestions, and are now the default method to select a panel visualization.
Expo SDK 55 Beta is now available
Learn about the changes, how to try it out, and how to give feedback.
Modal 1.3.1
- We've improved our experimental support for Python 3.14t (free-threaded Python) inside Modal containers. The container environment will now use the Python implementation of the Protobuf runtime rather than the incompatible upb implementation.
Cloudflare One, Access - Require Access protection for zones
You can now require Cloudflare Access protection for all hostnames in your account. When enabled, traffic to any hostname that does not have a matching Access application is automatically blocked. This deny-by-default approach prevents accidental exposure of internal resources to the public Internet.
Access - New granular API token permissions for Cloudflare Access
Three new API token permissions are available for Cloudflare Access, giving you finer-grained control when building automations and integrations: Access: Organizations Revoke , Grants the ability to revoke user sessions in a Zero Trust organization.
Rules - New cryptographic functions , encode_base64() and sha256()
Cloudflare Rulesets now includes encode_base64() and sha256() functions, enabling you to generate signed request headers directly in rule expressions. These functions support common patterns like constructing a canonical string from request attributes, computing a SHA256 digest, and Base64-encoding the result.
More from Cerebras InferenceEvery Cerebras Inference release →
API Version 2 Available for Testing API version 2 is now available for testing via the...
API Version 2 Available for Testing API version 2 is now available for testing via the X-Cerebras-Version-Patch header. This version introduces stricter validation and breaking changes.
Prioritize requests to the API to balance latency sensitivity and resource allocation across your workloads...
Prioritize requests to the API to balance latency sensitivity and resource allocation across your workloads with the new Service Tiers feature. Learn more in our Service Tiers guide.
Constrained Decoding Generally Available Constrained decoding has moved to General Availability (GA) with...
Constrained Decoding Generally Available Constrained decoding has moved to General Availability (GA) with expanded capabilities. Supported models Structured Outputs and Tool Calling with strict: true are now available on the following models: gpt-oss-120b llama-3.1-8b qwen-3-235b-a22b-instruct-2507 qwen-3-32b zai-glm-4.6 zai-glm-4.7 New capabilities...
Process large-scale inference workloads asynchronously with the new Batch API and Files API
Process large-scale inference workloads asynchronously with the new Batch API and Files API. Submit up to 50,000 requests in a single batch job and get 50% cost savings. Learn more in our Batch guide.
reasoning_effort="none" support for GLM 4.7 zai-glm-4.7 now accepts the reasoning_effort parameter
reasoning_effort="none" support for GLM 4.7 zai-glm-4.7 now accepts the reasoning_effort parameter. Reasoning is enabled by default , set reasoning_effort="none" to disable it for simple tasks where verbose reasoning blocks are unnecessary. The disable_reasoning parameter continues to be supported but will be deprecated soon.
Added support for parallel tool calling, allowing models to request multiple tool invocations simultaneously...
Added support for parallel tool calling, allowing models to request multiple tool invocations simultaneously in a single response. This reduces latency when handling queries that require multiple independent operations. Learn more in our Tool Calling guide.
Streaming Support Streaming is now supported for all models, including with structured output requests...
Streaming Support Streaming is now supported for all models, including with structured output requests (response_format, tools). Reasoning Format Controls Added reasoning_format parameter for reasoning models, giving you control over how reasoning text appears in responses.