AI.info
AI Tools Updates — Page 204
What changed in the AI tools this week — dated, from the makers’ own changelogs.
- HyperCLOVA X — HyperCLOVA X THINK: From seeds to forest
From seeds to forest.
- Nosto Search & Discovery — What’s New in Nosto: 3x Faster Search and Merchandising
3x Faster Search and Merchandising
- Hudl — Relive Your Season: Introducing Hudl Rewind
Introducing Hudl Rewind.
- Moshi — mlx-0.2.12
The Python package versions were bumped.
- Moshi — moshi-v0.2.11
The Python package versions were bumped.
- Deepdub GO — Deepdub Brings Voice API for AI Agents in AWS Marketplace's AI Agents and Tools Storefront
Brings Voice API for AI Agents to AWS Marketplace's AI Agents and Tools Storefront.
- Deepdub GO — Deepdub Announces Availability of Voice API for AI Agents in the New AWS Marketplace AI Agents and Tools Storefront
Announces Availability of Voice API for AI Agents in the AWS Marketplace AI Agents and Tools Storefront.
- Helicone — Improved Timezone Display for Requests
Requests now display in your local timezone automatically, with timezone indicators and date/time formatting.
- Rask AI — July 15, 2025 – More Reliable AI Voice Presets with New Provider
AI Voice Presets with New Provider.
- Rask AI — More Reliable AI Voice Presets with New Provider
AI Voice Presets with New Provider.
- Helicone — Country-Based Request Filtering
Added country code filtering, geographic analytics, and usage patterns in the requests and analytics dashboards.
- Waabi World — The ultimate driving test for AI: Mixed Reality Testing pushes the boundaries of AV safety
Mixed Reality Testing pushes the boundaries of AV safety.
- D-ID — Agents 2.0 Now Available
New agents created in Studio now use the Agents 2.0 spec, with a 16:9 chat interface, customizable personalities, and real-time analytics.
- HeadshotPro — Automate Your Headshot Workflow with Our Zapier Integration
HeadshotPro now integrates with Zapier to manage headshots for teams.
- Sonic — Introducing Ink: speech-to-text models for real-time conversation
Introducing Ink, a family of streaming speech-to-text models for developers building real-time voice applications.
- Moshi — moshi-v0.2.10
The Moshi version number was bumped.
- Solar Pro 2 — Solar Pro 2: Fluent. Reasoning. Frontier.
Solar Pro 2 supports multilingual processing, advanced reasoning, and real-world tool use.
- Rask AI — July 9, 2025 – Long Video Processing Fixed & Additional Minutes for Creator Pro
Video Processing Fixed & Minutes for Creator Pro.
- Frase — Introducing Terms: Keep AI on brand, word for word
Terms helps your team and your AI stay aligned on the exact words that do, or don’t, belong in your content.
- Rask AI — Long Video Processing Fixed & Additional Minutes for Creator Pro
Long Video Processing fixed and Additional Minutes for Creator Pro.
- GrowthBook — v4.0.0
Added an MCP server, safer rollout features, customizable decision criteria, and search filters.
- HHEM (Hughes Hallucination Evaluation Model) — update HHEM-2.1-Open demo link
- Moshi — mlx-0.2.11: Moshi mlx.0.2.11 (#333)
The MLX local-web example was fixed, ported to sphn 0.2, and the Moshi MLX dependency was bumped.
- Lindy — Autosave and drafts, Gemini 2.5 Flash Lite
Changes to your agents will now be autosaved as drafts; discard draft reverts your agent to the last saved version. Gemini 2.5 Flash-Lite was added.
- Coze — Model settings support deep thinking
Developers can enable or disable Deep Thinking to flexibly control the token consumption of the model during interactions.
- Moshi — moshi-v0.2.8
Moshi 0.2.8 was released.
- Alitu — Get even closer to that smooth radio voice with a new audio filter: Bandwidth Extension
We've added Bandwidth Extension; enable it in Show Settings > App Preferences > Audio Clean-up.
- Alitu — Reach listeners almost instantly with PodPing support
Alitu now supports PodPing, so episodes can get to listeners instantly after you hit Publish.
- Unmute — Kyutai TTS and Unmute now open-source
Announcing the open-source release of Kyutai TTS 1.6B and Unmute, with benchmarks and project details.
- Moshi — moshi-v0.2.7
Moshi 0.2.7 was released.
- Applitools Autonomous — Autonomous 2.2
Generate test steps, generate test data, build API requests, and create a shared visual baseline for each unique combination of test parameter values.
- Applitools Autonomous — We’re introducing smarter test step generation, effortless data creation, intuitive API testing, and seamless cross-environment validation.
Introduces LLM-generated test steps and test data, an API Step Builder, and a single baseline per parameter value.
- Readwise — AI Themed Reviews 🧠
Creating a Themed Review is now as easy as entering a natural language topic.
- Readwise — Audio Reviews 🎧
You can now listen to your Readwise Daily Reviews on the go.
- SideChef — SideChef Shoppable Recipes Takes the Spotlight on Yahoo Search
Shoppable Recipes takes the spotlight on Yahoo Search.
- Baidu Wenxin Assistant — Announcing the Open Source Release of the ERNIE 4.5 Model Family
Open source release of the ERNIE 4.5 model family.
- Arcee-SuperNova — Arcee-SuperNova-v1
A merged 70B model built from multiple training approaches, targeting general intelligence applications and mathematical reasoning.
- Chai-1 — Chai Discovery Unveils Chai-2 Breakthrough, Achieving Fully De Novo Antibody Design With AI
Achieving Fully De Novo Antibody Design With AI.
- Chai-1 — Introducing Chai-2: Zero-shot antibody design in a 24-well plate
Models achieve double digit hit rates in de novo antibody design.
- Arcee-SuperNova — Releasing Five Open-Weights Models
Announced the open-weights release of five language models.