AI.info
AI Tools Updates — Page 98
What changed in the AI tools this week — dated, from the makers’ own changelogs.
Auf Englisch
- Palantir Foundry — Fireworks models now available for US georestricted commercial enrollments
The following models are now available in AIP through Fireworks for US georestricted commercial enrollments.
- vLLM — v0.31.0rc3: [Model Runner V2] Support randomized dummy inputs (#58411)
Adds support for randomized dummy inputs to Model Runner V2.
- Retool Agents — Claude Opus 5.5 available in Retool
Claude Opus 5.5 is available through Anthropic and Amazon Bedrock for AI resource queries on Enterprise organizations with your own key.
- Retool Agents — Retool-provided AI models on the Team plan
Team plan organizations can use Retool-provided AI models for app building alongside the supported model catalog.
- Skyvern — Release v1.0.55
Run reliability is shown inline below analytics.
- Snowflake Data Cloud — Snowflake Data Clean Rooms updates
- Snowflake Data Cloud — Migration of gen 1 Openflow deployments and runtimes to gen 2 (General availability)
Migration of gen 1 Openflow deployments and runtimes to gen 2
- Consensus — October 1st, 2026 - New publisher partnerships with ASM and Rockefeller University Press
Consensus has partnered with ASM and Rockefeller University Press to bring biomedical full-text content into Consensus search and analysis.
- Crowdin — What’s new at Crowdin: September 2026
Crowdin Copilot can run scripts and build apps; AI Pipeline can filter ambiguous strings, and glossary and TM can cover every project.
- CodeRabbit — GitLab and Bitbucket Cloud automatic linking | Team
Automatic repository linking now supports GitLab, including self-managed instances, and Bitbucket Cloud.
- CodeRabbit — Author approval control
A setting can prevent pull request authors from approving or resolving their own pull requests.
- Unsloth — Command Palette + Desktop UI/UX
Unsloth Desktop adds a command palette, shareable GGUF run settings, clearer errors, faster Laya decisions, hosted Decision API support, and packed 4-bit LoRA training.
- TT-Metalium — v0.80.0-dev20261001
Disallowed 2x data format with GMPOOL and banned TT_/TTI_ macros outside tensix.
- Retool — Claude Opus 5.5 available in Retool
Claude Opus 5.5 is available in Retool through Anthropic and Amazon Bedrock for AI resource queries on Enterprise.
- Retool — Retool-provided AI models on the Team plan
Team plan organizations can use Retool-provided AI models for app building.
- Kilo Code — v7.8.3 (release)
Added document viewer copy controls, a manual Fetch Models button, and file suggestions from all folders in multi-root workspaces.
- Opik — 2.2.87
- Opik — 2.2.88-6876
Renders compare prompts as messages.
- Keploy — v3.6.84
Observe-only capture records the part of an HTTP response its client read.
- Tonic Structural — v1882
Bug fixes and other internal updates.
- Tonic Structural — v1881
On Privacy Hub, fixed an issue that caused an error when you clicked the magnifying glass icon to preview a column's data.
- Snowflake Cortex Analyst — Migration of gen 1 Openflow deployments and runtimes to gen 2 (General availability)
Migration of gen 1 Openflow deployments and runtimes to gen 2
- Snowflake Cortex Analyst — Snowflake Data Clean Rooms updates
- llama.cpp — b11317
Fixed llama-bench's verbosity filter to show GGML_LOG_ERROR.
- llama.cpp — b11319
- llama.cpp — b11321
Added an AOCL-BLAS quick start and labeled the device AOCL-BLAS.
- llama.cpp — b11322
Fixed a race condition that could desynchronize seqn from idx_read/write.
- llama.cpp — b11318
PLaMo-2 and PLaMo-3 now honor BOS/EOS settings from tokenizer_config.json during conversion and tokenization.
- llama.cpp — b11316
- llama.cpp — b11313
Re-enabled -sm tensor for qwen4exp.
- llama.cpp — b11320
Added a dedicated Harmony dialect handler for LLM-jp-4.1, selected with chat_format=llm-jp-harmony-v1.
- JetBrains AI Assistant — A New Agentic Experience: JetBrains Air in IDEs – EAP Now Open
JetBrains Air entered EAP for JetBrains IDEs as a Marketplace plugin and in 2026.3 EAP builds, with free Junie Lite runs.
- Everlaw Ediscovery Platform — New Harvey + Everlaw Integration Brings Massive Scale to Legal AI
A new integration connects Harvey with Everlaw’s governed evidence layer for legal AI workflows.
- Everlaw Ediscovery Platform — Everlaw Brings Frontier AI Closer to the Evidence with ChatGPT Plugin
Everlaw launched a ChatGPT Enterprise plugin connecting its governed evidence layer with ChatGPT Enterprise.
- Everlaw Ediscovery Platform — Everlaw and Thomson Reuters CoCounsel Legal Launch New Partnership
Everlaw and Thomson Reuters announced a forthcoming integration connecting matter facts with the law in AI tools.
- Everlaw Ediscovery Platform — Everlaw and Microsoft Copilot Announce New MCP Integration (Available Now)
An MCP integration brings Everlaw’s evidentiary system of record to Microsoft 365 through Copilot; available now.
- Fireworks AI — Serverless pricing update: DeepSeek V4.1 Flash
Serverless pricing for DeepSeek V4.1 Flash changes; dedicated deployment and Reserved Throughput pricing is unaffected.
- Genesis — v1.4.3
Improved GPU simulation speed and reduced memory usage; compilation is up to three times faster on Windows.
- CoreWeave Cloud — CoreWeave Delivers NVIDIA Vera Rubin NVL72 Performance at Production Scale, Starting With Cognition
CoreWeave maintains industry-leading AI cloud performance across multiple compute generations with its software platform
- CoreWeave Cloud — CoreWeave to Offer NVIDIA Vera, the First CPU Built for AI agents
CoreWeave expands compute portfolio to meet the needs of AI workloads.