Mini tools
ElevenLabs
163,386
MiniMax Speech
4,919
Lightning TTS
3,744
DeepL
35,260
ElevenLabs Dubbing
7,547
Stable Audio
4,510
Minimax Music
1,190
Lyria
6,097
Grok Imagine
73,253
Vdoo AI | Image to Video Generator
10,541
Vdoo AI | Reference to Video
16,085
H2O.ai
5,980
Arcee AI
108
Aster Lab
3,561
Devin Review
394
Decagon
77
Nano Banana
10,085
Vdoo AI | Image to Image Generator
8,845
Vdoo AI | Image Editor
2,673
JobDescriptionGenerator
1,149
Ai2
Follow
Visit website
Follow
Visit website
Ai2 is a 501(c)(3) non-profit AI research institute founded in 2014 by the late Paul Allen (Microsoft co-founder), dedicated to conducting high-impact, open AI research and engineering for the common good, including open language models (OLMo), scientific AI tools (Semantic Scholar, Asta), environmental AI platforms, and embodied robotics research.
AI Native
Yes
Number of tools
0
Number of models
11
Number of employees
313
Valuation
$500MAI
Models
-
4B vision-language model that forecasts 3D point trajectories from a single RGB observation frame, 2D query points, and a natural-language action instruction. Built on Molmo 2, it predicts where surface points on objects will move in 3D camera-frame space (in meters) over 32 future timesteps. Best used when only one observation keyframe is available.NewMultimodalReleased 24d ago
-
MolmoAct 2 is Ai2’s open robotics foundation model for real-world robot manipulation. It is an Action Reasoning Model that reasons about 3D structure before acting, supports bimanual manipulation out of the box, and is designed for zero-shot deployment and post-training adaptation across household, lab, and tabletop tasks.NewMultimodalReleased 2mo ago
-
F_pre_8kv_8k_14k_SWA is an AllenAI OlmPool architectural-variation model for studying long-context extension. It is a 7B to 8B class language model trained to 150B tokens total, including 140B pretraining and 10B context extension, and is meant for research on architecture choices rather than polished instruction-following use.NewTextReleased 2mo ago
-
WildDet3D is Ai2’s open model for monocular 3D object detection from a single RGB image. It predicts full 3D bounding boxes, including position, size, and orientation in metric coordinates, and supports flexible prompts such as text queries, point clicks, and 2D boxes. It is designed for open-world spatial understanding across varied cameras and can also use optional depth signals when available.3dReleased 3mo ago
-
SERA-32B-GA is a 32B Qwen 3 based coding agent variant trained on GLM-4.5-Air trajectories, delivering about 46.6% SWE-bench Verified performance as a slightly smaller-teacher alternative to SERA-32B.TextReleased 5mo ago
-
SERA-32B is Ai2’s flagship 32B Open Coding Agent, a Qwen 3-32B based repository-level coding model trained with Soft Verified Generation that reaches about 49.5% on SWE-bench Verified, rivaling much larger proprietary systems.TextReleased 5mo ago
-
SERA-8B-GA is an 8B Qwen 3 based coding agent variant trained from GLM-4.5-Air generated trajectories, matching SERA-8B’s 31.7% SWE-bench Verified score while providing an alternative teacher setup for research.TextReleased 5mo ago
-
SERA-8B is Ai2’s 8B-parameter open-source coding agent, fine-tuned from Qwen 3-8B with Soft Verified Generation to solve repository-level bugs, reaching about 31.7% resolve rate on SWE-bench Verified at 32K context.TextReleased 5mo ago
-
Molmo 2 is AI2’s open 4B/7B/8B multimodal model for images and video, delivering state-of-the-art grounded video QA, pointing and tracking that return coordinates and timestamps for events instead of text-only answers.TextReleased 6mo ago
-
OlmoEarth is Ai2/AllenAI’s foundation-model collection for remote-sensing tasks using satellite and Earth-observation imagery.MultimodalReleased 8mo ago
-
olmOCR is AllenAI’s open-source document recognition pipeline and model family that converts PDFs and images into clean text, preserving reading order, tables, equations, and handwriting.ImageReleased 8mo ago
