---
title: "AI models released in July 2025"
type: "release-archive"
canonical_url: "https://theaiengineering.com/releases/2025/07"
markdown_url: "https://theaiengineering.com/releases/2025/07.md"
author: "The AI Engineering Editors"
publisher: "The AI Engineering"
updated: "2025-07-31"
citation: "The AI Engineering Editors. \"AI models released in July 2025.\" The AI Engineering, updated Jul 31, 2025. https://theaiengineering.com/releases/2025/07"
license: "CC BY 4.0"
license_url: "https://creativecommons.org/licenses/by/4.0/"
data_csv: "https://theaiengineering.com/releases.csv"
data_json: "https://theaiengineering.com/releases.json"
---

# AI models released in July 2025

34 releases from 20 labs, 26 with open weights. Dates are the labs' official announcement dates.

## July 2025

### Command A Vision

Cohere's first Command model that understands images, built for enterprise visual tasks such as charts, diagrams, PDFs, slides and real-world scenes.

- Date: 2025-07-31
- Lab: Cohere
- Type: Multimodal
- Weights: Open weights
- Modalities: text, image → text
- Context window: 128,000 tokens
- Parameters: 112B
- License: CC-BY-NC-4.0
- Available at launch: API, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#command-a-vision

Highlights:

- Deploys privately on two A100 GPUs or one H100 with 4-bit quantization
- Keeps Command A's RAG with citations
- Research weights released as command-a-vision-07-2025

Source: [Command A Vision: Multimodal AI for Business | Cohere](https://cohere.com/blog/command-a-vision)

### FLUX.1 Krea [dev]

Open-weight text-to-image model developed with Krea AI as the open version of Krea 1, trained to avoid the oversaturated AI look and produce more realistic, diverse images.

- Date: 2025-07-31
- Lab: Black Forest Labs
- Type: Image
- Weights: Open weights
- Modalities: text → image
- License: FLUX.1 [dev] Non-Commercial License
- Available at launch: Open weights, Cloud platforms
- Permalink: https://theaiengineering.com/releases/2025/07#flux-1-krea-dev

Highlights:

- Architecturally compatible with the FLUX.1 [dev] ecosystem
- Weights on Hugging Face, commercial licenses via the BFL Licensing Portal
- Positioned as a base model for downstream fine-tuning

Source: [FLUX.1 Krea \[dev\]: An 'Opinionated' Text-to-Image Model](https://bfl.ai/blog/flux-1-krea-dev)

### Codestral 25.08

Update to the Codestral fill-in-the-middle model, announced as part of Mistral's coding stack for enterprises with more accepted completions and fewer runaway generations.

- Date: 2025-07-30
- Lab: Mistral AI
- Type: Coding
- Weights: Proprietary
- Modalities: text → text
- Context window: 128,000 tokens
- Available at launch: API, Enterprise
- Permalink: https://theaiengineering.com/releases/2025/07#codestral-2508

Highlights:

- Deployable in cloud, VPC or on-premises environments
- Improved chat-mode instruction following
- API model ID codestral-2508

Source: [Announcing Codestral 25.08 and the Complete Mistral Coding Stack for Enterprise | Mistral AI](https://mistral.ai/news/codestral-25-08)

### GLM-4.5

GLM-4.5 is Z.ai's open-weights foundation model series for agents, unifying reasoning, coding, and agentic abilities. It comes in GLM-4.5 and the more compact GLM-4.5-Air, both with thinking and non-thinking modes.

- Date: 2025-07-28
- Lab: Z.ai
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- Context window: 128,000 tokens
- Parameters: MoE: 355B total, 32B active; Air 106B total, 12B active
- License: MIT
- Available at launch: API, Consumer app, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#glm-4-5

Highlights:

- Pre-trained on 15T tokens of general data, then on code, reasoning, and agent data
- 128K context length
- Weights on Hugging Face and ModelScope under the MIT license

Source: [GLM-4.5: Reasoning, Coding, and Agentic Abilities](https://z.ai/blog/glm-4.5)

### Wan2.2

Wan2.2 is the open video generation family from Alibaba's Wan team, adding mixture-of-experts text-to-video and image-to-video models and a compact 5B text-and-image-to-video model.

- Date: 2025-07-28
- Lab: Alibaba Qwen
- Type: Video
- Weights: Open weights
- Modalities: text, image → video
- Parameters: MoE: A14B (T2V, I2V); 5B (TI2V)
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#wan2-2

Highlights:

- T2V-A14B and I2V-A14B support 480P and 720P
- TI2V-5B uses a high-compression VAE and supports 720P
- Integrated into Diffusers and ComfyUI at launch; Apache 2.0 license

Source: [Wan2.2 (GitHub)](https://github.com/Wan-Video/Wan2.2)

### HunyuanWorld 1.0

HunyuanWorld 1.0 is Tencent's open model for generating immersive, explorable 360-degree 3D worlds from text or images, combining panoramic proxy generation, semantic layering, and hierarchical 3D reconstruction.

- Date: 2025-07-26
- Lab: Tencent
- Type: Image
- Weights: Open weights
- Modalities: text, image → 3d
- License: Tencent HunyuanWorld-1.0 Community License Agreement
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#hunyuanworld-1-0

Highlights:

- Tencent describes it as the first open-source, simulation-capable immersive 3D world generation model
- Quantized HunyuanWorld-1.0-lite for consumer GPUs added August 15, 2025

Source: [GitHub - Tencent-Hunyuan/HunyuanWorld-1.0: Generating Immersive, Explorable, and Interactive 3D Worlds from Words or Pixels with Hunyuan3D World Model](https://github.com/Tencent-Hunyuan/HunyuanWorld-1.0)

### Aleph

In-context video model for multi-task editing that adds, removes and transforms objects, generates new angles of a scene and modifies style and lighting on an input video.

- Date: 2025-07-25
- Lab: Runway
- Type: Video
- Weights: Proprietary
- Modalities: text, image, video → video
- Available at launch: Consumer app
- Permalink: https://theaiengineering.com/releases/2025/07#runway-aleph

Highlights:

- Available to all paid plans from July 31, 2025
- Generates new camera angles from existing footage
- Succeeded by Aleph 2.0 in May 2026

Source: [Introducing Runway Aleph](https://runway.com/research/introducing-runway-aleph)

### Llama-3.3-Nemotron-Super-49B-v1.5

Significantly upgraded version of Llama-3.3-Nemotron-Super-49B, a reasoning model NVIDIA derived from Meta's Llama-3.3-70B-Instruct using neural architecture search to cut memory needs while keeping accuracy.

- Date: 2025-07-25
- Lab: NVIDIA
- Type: Reasoning
- Weights: Open weights
- Modalities: text → text
- Parameters: 49B
- License: NVIDIA Open Model License
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#llama-3-3-nemotron-super-49b-v1-5

Highlights:

- Trained between November 2024 and July 2025
- Post-trained for reasoning, chat preferences and agentic tasks such as RAG and tool calling
- Part of the Llama Nemotron collection

Source: [nvidia/Llama-3_3-Nemotron-Super-49B-v1_5 · Hugging Face](https://huggingface.co/nvidia/Llama-3_3-Nemotron-Super-49B-v1_5)

### Qwen3-235B-A22B-Thinking-2507

Qwen3-235B-A22B-Thinking-2507 is an updated thinking-mode version of Qwen3-235B-A22B, released as a separate reasoning model in the Qwen3-2507 series, which split the hybrid Qwen3 model into dedicated instruct and thinking models.

- Date: 2025-07-25
- Lab: Alibaba Qwen
- Type: Reasoning
- Weights: Open weights
- Modalities: text → text
- Context window: 262,144 tokens
- Parameters: MoE: 235B total, 22B active
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#qwen3-thinking-2507

Highlights:

- Native 262,144-token context
- Qwen3-30B-A3B-Thinking-2507 followed on July 31 and Qwen3-4B-Thinking-2507 on August 6, 2025
- Weights under the Apache 2.0 license

Source: [Qwen/Qwen3-235B-A22B-Thinking-2507 (Hugging Face)](https://huggingface.co/Qwen/Qwen3-235B-A22B-Thinking-2507)

### Step 3

Step 3 is StepFun's multimodal reasoning model built on a mixture-of-experts architecture and co-designed with Multi-Matrix Factorization Attention and Attention-FFN Disaggregation to reduce decoding costs.

- Date: 2025-07-25
- Lab: StepFun
- Type: Multimodal
- Weights: Open weights
- Modalities: text, image → text
- Context window: 65,536 tokens
- Parameters: MoE: 321B total, 38B active
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#step-3

Highlights:

- 321B total parameters (316B language model) with 38B active per token
- Maximum context length of 65,536 tokens
- Model weights were uploaded to Hugging Face on July 31, 2025 under Apache-2.0

Source: [stepfun-ai/step3 · Hugging Face](https://huggingface.co/stepfun-ai/step3)

### Qwen-MT

Qwen-MT (qwen-mt-turbo) is Qwen's translation model on the Qwen API, built on Qwen3 and trained on trillions of multilingual and translation tokens. It uses a lightweight mixture-of-experts architecture for low latency.

- Date: 2025-07-24
- Lab: Alibaba Qwen
- Type: LLM
- Weights: Proprietary
- Modalities: text → text
- Available at launch: API
- Permalink: https://theaiengineering.com/releases/2025/07#qwen-mt

Highlights:

- Translates across 92 major languages and dialects
- Supports terminology intervention, domain prompts, and translation memory
- Available through Alibaba Cloud Model Studio

Source: [Qwen-MT: Where Speed Meets Smart Translation](https://qwenlm.github.io/blog/qwen-mt/)

### Seed LiveInterpret 2.0

Seed LiveInterpret 2.0 is an end-to-end simultaneous interpretation model for Chinese-to-English and English-to-Chinese speech, built on a full-duplex framework with real-time voice cloning so translations keep the speaker's voice.

- Date: 2025-07-24
- Lab: ByteDance Seed
- Type: Speech
- Weights: Proprietary
- Modalities: audio → audio, text
- Available at launch: API
- Permalink: https://theaiengineering.com/releases/2025/07#seed-liveinterpret-2-0

Highlights:

- Covers Chinese-to-English and English-to-Chinese interpretation
- Clones the speaker's voice in real time
- Offered on Volcano Engine as Doubao Simultaneous Interpretation 2.0
- Ola Friend earbuds announced as the first hardware to adopt it

Source: [Seed LiveInterpret 2.0 Released: An End-to-End Simultaneous Interpretation Model Featuring Ultra-High Accuracy Close to Human Interpreters, Low Latency of 3 Seconds, and Real-Time Voice Cloning](https://seed.bytedance.com/en/blog/seed-liveinterpret-2-0-released-an-end-to-end-simultaneous-interpretation-model-featuring-ultra-high-accuracy-close-to-human-interpreters-low-latency-of-3-seconds-and-real-time-voice-cloning)

### Qwen3-Coder

Qwen3-Coder is Alibaba's agentic coding model family, led by Qwen3-Coder-480B-A35B-Instruct. Alibaba also open-sourced Qwen Code, a command-line tool for agentic coding in the terminal, alongside it.

- Date: 2025-07-22
- Lab: Alibaba Qwen
- Type: Coding
- Weights: Open weights
- Modalities: text → text
- Context window: 262,144 tokens
- Parameters: MoE: 480B total, 35B active
- License: Apache-2.0
- Available at launch: API, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#qwen3-coder

Highlights:

- Pre-trained on 7.5T tokens with a 70% code ratio
- Native 256K context, extendable to 1M with YaRN
- Qwen3-Coder-30B-A3B-Instruct followed on July 31, 2025
- Weights under the Apache 2.0 license

Source: [Qwen3-Coder: Agentic Coding in the World](https://qwenlm.github.io/blog/qwen3-coder/)

### Qwen3-235B-A22B-Instruct-2507

Qwen3-235B-A22B-Instruct-2507 is an updated non-thinking version of Qwen3-235B-A22B with significant improvements and 256K-token long-context understanding. It began the Qwen3-2507 series of separate instruct and thinking models.

- Date: 2025-07-21
- Lab: Alibaba Qwen
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- Context window: 262,144 tokens
- Parameters: MoE: 235B total, 22B active
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#qwen3-instruct-2507

Highlights:

- Native 262,144-token context; 1M-token support added August 8, 2025
- Qwen3-30B-A3B-Instruct-2507 followed on July 30 and Qwen3-4B-Instruct-2507 on August 6, 2025
- Weights under the Apache 2.0 license

Source: [Qwen3 (GitHub)](https://github.com/QwenLM/Qwen3)

### Seed-X

Seed-X is an open-weights family of 7B multilingual translation models from ByteDance Seed, comprising instruction, reinforcement-learning, and reward models, supporting translation among 28 languages.

- Date: 2025-07-18
- Lab: ByteDance Seed
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- Parameters: 7B
- License: OpenMDW-1.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#seed-x

Highlights:

- Seed-X-Instruct, Seed-X-PPO, and Seed-X-RM released on Hugging Face
- Uses a Mistral-style architecture
- Quantized 8-bit and 4-bit PPO models followed on July 28, 2025

Source: [GitHub - ByteDance-Seed/Seed-X-7B](https://github.com/ByteDance-Seed/Seed-X-7B)

### Act-Two

Next-generation motion capture model succeeding Act-One, with major gains in generation quality and support for head, face, body and hand tracking for character performances.

- Date: 2025-07-17
- Lab: Runway
- Type: Video
- Weights: Proprietary
- Modalities: video, image → video
- Available at launch: Consumer app
- Permalink: https://theaiengineering.com/releases/2025/07#runway-act-two

Highlights:

- Head, face, body and hand tracking
- Available on paid plans
- Voice changing added to the Act-Two interface in August 2025

Source: [Product Updates & Changelog | Runway AI](https://runway.com/changelog)

### Canary-Qwen-2.5B

English speech recognition model built as a speech-augmented language model from a FastConformer Canary encoder and the Qwen3-1.7B language model.

- Date: 2025-07-17
- Lab: NVIDIA
- Type: Speech
- Weights: Open weights
- Modalities: audio → text
- Parameters: 2.5B
- License: CC-BY-4.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#canary-qwen-2-5b

Highlights:

- Combines canary-1b-flash with Qwen3-1.7B through LoRA
- Supports English speech-to-text
- Released under CC BY 4.0

Source: [nvidia/canary-qwen-2.5b · Hugging Face](https://huggingface.co/nvidia/canary-qwen-2.5b)

### LTX-Video 0.9.8

LTX-Video update adding long-shot generation of up to 60 seconds of video, with new 13B and 2B distilled models from a shared base and an IC-LoRA detailer model.

- Date: 2025-07-16
- Lab: Lightricks
- Type: Video
- Weights: Open weights
- Modalities: text, image → video
- Parameters: 13B and 2B
- License: LTXV Open Weights License 0.X
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#ltx-video-0-9-8

Highlights:

- Up to 60 seconds of video
- 13B and 2B distilled variants usable in one multiscale pipeline
- Compatible with the official IC-LoRA control models

Source: [Lightricks/LTX-Video](https://github.com/Lightricks/LTX-Video)

### OpenReasoning-Nemotron

Reasoning models derived from Qwen2.5 base models and post-trained for math, code and science solutions, released in four sizes under CC BY 4.0.

- Date: 2025-07-16
- Lab: NVIDIA
- Type: Reasoning
- Weights: Open weights
- Modalities: text → text
- Parameters: 1.5B, 7B, 14B and 32B
- License: CC-BY-4.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#openreasoning-nemotron

Highlights:

- 32B model developed from Qwen2.5-32B
- Covers math, code and science reasoning
- Released under CC BY 4.0

Source: [nvidia/OpenReasoning-Nemotron-32B · Hugging Face](https://huggingface.co/nvidia/OpenReasoning-Nemotron-32B)

### CogVideoX-3

CogVideoX-3 is Z.ai's upgraded video generation model on its API, supporting text-to-video and image-to-video. It adds first-and-last-frame generation and improves clarity and the smoothness of large subject motion.

- Date: 2025-07-15
- Lab: Z.ai
- Type: Video
- Weights: Proprietary
- Modalities: text, image → video
- Available at launch: API
- Permalink: https://theaiengineering.com/releases/2025/07#cogvideox-3

Highlights:

- Adds start and end frame control
- Improved realistic and 3D-style scenes
- Available through the BigModel API platform

Source: [Model and product release records (BigModel docs)](https://docs.bigmodel.cn/cn/update/new-releases)

### Voxtral Small and Voxtral Mini

Mistral's first audio models, open-weight speech understanding models that combine transcription with question answering and summarization, in a 24B production size and a 3B edge size.

- Date: 2025-07-15
- Lab: Mistral AI
- Type: Speech
- Weights: Open weights
- Modalities: text, audio → text
- Context window: 32,000 tokens
- Parameters: 24B and 3B
- License: Apache-2.0
- Available at launch: Open weights, API
- Permalink: https://theaiengineering.com/releases/2025/07#voxtral

Highlights:

- Handles audio up to 30 minutes for transcription or 40 minutes for understanding
- The API also offers Voxtral Mini Transcribe for transcription
- Released under Apache 2.0

Source: [Voxtral | Mistral AI](https://mistral.ai/news/voxtral)

### Kimi K2

Kimi K2 is Moonshot AI's open-weights mixture-of-experts model optimized for agentic tasks, released as Kimi-K2-Base for fine-tuning and Kimi-K2-Instruct for chat and agent use. Moonshot trained it with the MuonClip optimizer.

- Date: 2025-07-11
- Lab: Moonshot AI
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- Context window: 128,000 tokens
- Parameters: MoE: 1T total, 32B active
- License: Modified MIT License
- Available at launch: API, Consumer app, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#kimi-k2

Highlights:

- 1T total parameters with 32B activated
- 128K context length at launch
- Offered on the Kimi Open Platform as kimi-k2-0711-preview
- Weights and code under the Modified MIT License

Source: [Kimi K2: Open Agentic Intelligence](https://www.kimi.com/en/blog/kimi-k2)

### Devstral Medium

Larger agentic coding model released through the API alongside Devstral Small 1.1, built with All Hands AI for software engineering agents.

- Date: 2025-07-10
- Lab: Mistral AI
- Type: Coding
- Weights: Proprietary
- Modalities: text → text
- Context window: 128,000 tokens
- Available at launch: API, Enterprise
- Permalink: https://theaiengineering.com/releases/2025/07#devstral-medium

Highlights:

- API model ID devstral-medium-2507
- Available for on-premises deployment and fine-tuning for enterprises
- Deprecated February 27, 2026

Source: [Upgrading agentic coding capabilities with the new Devstral models | Mistral AI](https://mistral.ai/news/devstral-2507)

### Devstral Small 1.1

Upgrade to Devstral Small that keeps the 24B architecture while improving generalization across prompts and agent scaffolds, built with All Hands AI.

- Date: 2025-07-10
- Lab: Mistral AI
- Type: Coding
- Weights: Open weights
- Modalities: text → text
- Context window: 128,000 tokens
- Parameters: 24B
- License: Apache-2.0
- Available at launch: Open weights, API
- Permalink: https://theaiengineering.com/releases/2025/07#devstral-small-1-1

Highlights:

- Supports both Mistral function calling and XML formats
- Released under Apache 2.0
- API model ID devstral-small-2507

Source: [Upgrading agentic coding capabilities with the new Devstral models | Mistral AI](https://mistral.ai/news/devstral-2507)

### LFM2 (350M, 700M, 1.2B)

Second-generation Liquid Foundation Models using a new hybrid architecture of gated short convolutions and grouped query attention, released as open-weight 350M, 700M, and 1.2B dense checkpoints designed for fast on-device inference.

- Date: 2025-07-10
- Lab: Liquid AI
- Type: Small model
- Weights: Open weights
- Modalities: text → text
- Context window: 32,768 tokens
- Parameters: 350M: 354M; 700M: 742M; 1.2B: 1.17B
- License: LFM Open License v1.0
- Available at launch: API, Consumer app, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#lfm2

Highlights:

- Architecture: 10 double-gated short convolution blocks and 6 grouped query attention blocks
- Trained on 10T tokens with LFM1-7B as a knowledge-distillation teacher
- Weights on Hugging Face; also on Liquid Playground and OpenRouter
- LFM Open License v1.0 allows free commercial use below a company revenue threshold

Source: [Introducing LFM2: The Fastest On-Device Foundation Models on the Market](https://www.liquid.ai/blog/liquid-foundation-models-v2-our-second-series-of-generative-ai-models)

### Reka Flash 3.1

An update to Reka Flash 3 from an upgraded reinforcement learning stack, aimed at coding and as a base for agentic fine-tuning, released with a 3.5-bit Reka Quant version and the Reka Quant quantization library.

- Date: 2025-07-10
- Lab: Reka
- Type: Reasoning
- Weights: Open weights
- Modalities: text → text
- Parameters: 21B
- License: Apache-2.0
- Available at launch: API, Consumer app, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#reka-flash-3-1

Highlights:

- API model name reka-flash-3.1; also available in the Reka Playground
- 3.5-bit Q3_K_S GGUF build quantized with Reka Quant for llama.cpp
- A customized version orchestrates Reka Research; a multimodal version underpins Reka Vision
- Weights on Hugging Face under Apache-2.0

Source: [Reka Flash 3.1 and Reka Quant](https://reka.ai/news/reka-flash-3-1-and-reka-quant)

### FlexOlmo

Language model training paradigm in which each data owner branches from a shared public model and trains an expert locally on its own data, letting owners opt in or out later.

- Date: 2025-07-09
- Lab: Ai2
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#flexolmo

Highlights:

- Enables collaboration without pooling sensitive data
- Released as FlexOlmo-7x7B-1T on Hugging Face
- Released under Apache 2.0

Source: [Introducing FlexOlmo: a new paradigm for language model training and data collaboration | Ai2](https://allenai.org/blog/flexolmo)

### Grok 4 and Grok 4 Heavy

Grok 4 is xAI's flagship trained with reinforcement learning at pretraining scale on the 200,000-GPU Colossus cluster, with native tool use and real-time search. Grok 4 Heavy adds parallel test-time compute.

- Date: 2025-07-09
- Lab: xAI
- Type: LLM
- Weights: Proprietary
- Modalities: text, image → text
- Context window: 256,000 tokens
- Available at launch: Consumer app, API
- Permalink: https://theaiengineering.com/releases/2025/07#grok-4

Highlights:

- API offers a 256,000-token context window with text and vision understanding
- Available to SuperGrok and Premium+ subscribers and through the xAI API
- Grok 4 Heavy offered through a new SuperGrok Heavy tier
- API model grok-4-0709 retired May 15, 2026

Source: [Grok 4 | SpaceXSI](https://x.ai/news/grok-4)

### MedGemma 27B Multimodal

MedGemma 27B Multimodal extends Google's open medical model collection with a 27B model that accepts images and text, adding longitudinal electronic health record interpretation. It was released together with the MedSigLIP medical image encoder.

- Date: 2025-07-09
- Lab: Google
- Type: Multimodal
- Weights: Open weights
- Modalities: text, image → text
- Parameters: 27B
- License: Health AI Developer Foundations Terms of Use
- Available at launch: Open weights, Cloud platforms
- Permalink: https://theaiengineering.com/releases/2025/07#medgemma-27b-multimodal

Highlights:

- Complements the earlier MedGemma 4B Multimodal and 27B text models
- MedSigLIP is a roughly 400M-parameter SigLIP-based encoder for classification and retrieval
- All MedGemma models run on a single GPU

Source: [MedGemma: Our most capable open models for health AI development](https://research.google/blog/medgemma-our-most-capable-open-models-for-health-ai-development/)

### Phi-4-mini-flash-reasoning

Phi-4-mini-flash-reasoning is a 3.8B open Microsoft model for math reasoning on edge and mobile devices, built on the new SambaY decoder-hybrid-decoder architecture that combines Mamba, sliding window attention and gated memory units.

- Date: 2025-07-09
- Lab: Microsoft
- Type: Reasoning
- Weights: Open weights
- Modalities: text → text
- Context window: 64,000 tokens
- Parameters: 3.8B
- License: MIT
- Available at launch: Open weights, API, Cloud platforms
- Permalink: https://theaiengineering.com/releases/2025/07#phi-4-mini-flash-reasoning

Highlights:

- Uses the SambaY architecture with a Gated Memory Unit
- Supports a 64K token context length
- Available on Azure AI Foundry, NVIDIA API Catalog and Hugging Face

Source: [Reasoning reimagined: Introducing Phi-4-mini-flash-reasoning](https://azure.microsoft.com/en-us/blog/reasoning-reimagined-introducing-phi-4-mini-flash-reasoning/)

### T5Gemma

T5Gemma is a collection of encoder-decoder language models created by adapting pretrained Gemma 2 decoder-only models and further training them with UL2 or PrefixLM objectives. It includes balanced and unbalanced encoder-decoder pairings.

- Date: 2025-07-09
- Lab: Google
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- Parameters: 2B-2B, 9B-9B, 9B-2B, plus T5-sized Small to XL
- License: Gemma Terms of Use
- Available at launch: Open weights, Cloud platforms
- Permalink: https://theaiengineering.com/releases/2025/07#t5gemma

Highlights:

- Gemma 2-based configurations include 2B-2B, 9B-9B and an unbalanced 9B-2B
- Pretrained and instruction-tuned checkpoints
- Weights on Hugging Face and Kaggle; inference on Vertex AI

Source: [T5Gemma: A new collection of encoder-decoder Gemma models](https://developers.googleblog.com/en/t5gemma/)

### SmolLM3

Fully open 3B model with dual-mode reasoning, six languages and long context, released with an engineering blueprint covering its architecture and data mixtures.

- Date: 2025-07-08
- Lab: Hugging Face
- Type: Small model
- Weights: Open weights
- Modalities: text → text
- Context window: 128,000 tokens
- Parameters: 3B
- License: Apache-2.0
- Available at launch: Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#smollm3

Highlights:

- Trained on 11 trillion tokens
- Trained at 64K context and extends to 128K with YaRN
- Supports English, French, Spanish, German, Italian and Portuguese

Source: [SmolLM3: smol, multilingual, long-context reasoner](https://huggingface.co/blog/smollm3)

### Jamba 1.7

Update to Jamba Large and Jamba Mini with better grounding for more complete and accurate answers, plus optimized configurations and quantization.

- Date: 2025-07-03
- Lab: AI21 Labs
- Type: LLM
- Weights: Open weights
- Modalities: text → text
- License: Jamba Open Model License
- Available at launch: Open weights, API, Enterprise
- Permalink: https://theaiengineering.com/releases/2025/07#jamba-1-7

Highlights:

- Available for self-hosted deployment
- Weights on Hugging Face as Jamba Large 1.7 and Jamba Mini 1.7
- Suited to question answering and instruction following

Source: [Changelog - AI21 Labs](https://docs.ai21.com/changelog)

### GLM-4.1V-9B-Thinking

GLM-4.1V-9B-Thinking is Z.ai's open-weights vision-language reasoning model, built on GLM-4-9B-0414. It targets video understanding, image question answering, chart reading, and GUI operation with step-by-step reasoning.

- Date: 2025-07-01
- Lab: Z.ai
- Type: Multimodal
- Weights: Open weights
- Modalities: text, image, video → text
- Context window: 64,000 tokens
- Parameters: 9B
- License: MIT
- Available at launch: API, Open weights
- Permalink: https://theaiengineering.com/releases/2025/07#glm-4-1v-thinking

Highlights:

- Released with the GLM-4.1V-9B-Base model
- 64K context length support
- Weights under the MIT license; API launched July 2, 2025

Source: [GLM-V (GitHub)](https://github.com/zai-org/GLM-V)

## Cite and reuse this data

These release records are free to reuse under CC BY 4.0, including commercially. Credit The AI Engineering and link to this page's URL. Dates are the labs' official announcement dates, and every record links to its source. List fields in the CSV are separated by " | ".

The AI Engineering Editors. "AI models released in July 2025." The AI Engineering, updated Jul 31, 2025. https://theaiengineering.com/releases/2025/07

HTML link:

```html
<a href="https://theaiengineering.com/releases/2025/07">AI models released in July 2025</a> (The AI Engineering)
```

- Title: AI models released in July 2025
- Author: The AI Engineering Editors
- Publisher: The AI Engineering
- Updated: 2025-07-31
- URL: https://theaiengineering.com/releases/2025/07
- Markdown: https://theaiengineering.com/releases/2025/07.md
- License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)
- Full dataset as CSV: https://theaiengineering.com/releases.csv
- Full dataset as JSON: https://theaiengineering.com/releases.json
