Week 43, 2026 · Oct 19 to Oct 25, 2026

New in week 43

Tools first seen in the catalog this week, newest first. Also as an Atom feed (or just models).

All types Connectors Skills Plugins Harnesses & SDKs Agents No-code builders Models Agent infrastructure Observability & evals Identity, auth & payments Templates Protocols & benchmarks Artifacts

Any grade Security A Security B Security C Security D Security FSecurity grades are being rolled out; most listings are not graded yet.

695 listings first seen Oct 19 to Oct 25, 2026

0.395score
Llama-3.1-405B-Instruct
meta-llama
Open-weights 405.85B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.398score
Llama-3.1-70B
meta-llama
Open-weights 70.55B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.401score
Llama-3.1-70B-Instruct
meta-llama
Open-weights 70.55B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.462score
Llama-3.1-8B
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.466score
Llama-3.1-8B-Instruct
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.462score
Llama-3.2-1B
meta-llama
Open-weights 1.24B-parameter text model from meta-llama on Hugging Face (llama3.2).
ModelFFreshness: FOpen weights
0.462score
Llama-3.2-3B
meta-llama
Open-weights 3.21B-parameter text model from meta-llama on Hugging Face (llama3.2).
ModelFFreshness: FOpen weights
0.401score
Llama-Guard-3-8B
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.398score
Meta-Llama-3-70B
meta-llama
Open-weights 70.55B-parameter text model from meta-llama on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.401score
Meta-Llama-3-70B-Instruct
meta-llama
Open-weights 70.55B-parameter text model from meta-llama on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.462score
Meta-Llama-3-8B
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.462score
Meta-Llama-3-8B-Instruct
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.392score
Meta-Llama-Guard-2-8B
meta-llama
Open-weights 8.03B-parameter text model from meta-llama on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.327score
biogpt
microsoft
Open-weights text model from microsoft on Hugging Face (mit).
ModelFFreshness: FOpen weights
0.364score
bitnet-b1.58-2B-4T
microsoft
Open-weights 0.85B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.330score
DialoGPT-medium
microsoft
Open-weights text model from microsoft on Hugging Face (mit).
ModelFFreshness: FOpen weights
0.339score
MAI-DS-R1
microsoft
Open-weights 671.03B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.326score
Orca-2-13b
microsoft
Open-weights 13B-parameter text model from microsoft on Hugging Face (microsoft-research-license).
ModelFFreshness: FOpen weights
0.364score
phi-1_5
microsoft
Open-weights 1.42B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.423score
phi-2
microsoft
Open-weights 2.78B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.364score
Phi-3-medium-128k-instruct
microsoft
Open-weights 13.96B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.423score
Phi-3-mini-128k-instruct
microsoft
Open-weights 3.82B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.423score
Phi-3-mini-4k-instruct
microsoft
Open-weights 3.82B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.364score
Phi-3-vision-128k-instruct
microsoft
Open-weights 4.15B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.423score
Phi-3.5-mini-instruct
microsoft
Open-weights 3.82B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.364score
Phi-3.5-MoE-instruct
microsoft
Open-weights 41.87B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.423score
Phi-4-mini-instruct
microsoft
Open-weights 3.84B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.359score
Phi-4-reasoning-plus
microsoft
Open-weights 14.66B-parameter text model from microsoft on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.468score
Sentiment-Analysis-7B
microsoft
Open-weights 7.24B-parameter text model from microsoft on Hugging Face (mit).
ModelAFreshness: AOpen weights
0.414score
UserLM-8b
microsoft
Open-weights 8.03B-parameter text model from microsoft on Hugging Face (mit).
ModelCFreshness: COpen weights
0.488score
UserRepGen-4B
microsoft
Open-weights 4.54B-parameter text model from microsoft on Hugging Face (mit).
ModelAFreshness: AOpen weights
0.367score
MiniMax-M1-80k
MiniMaxAI
Open-weights 456.09B-parameter text model from MiniMaxAI on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.462score
Mistral-7B-Instruct-v0.1
mistralai
Open-weights 7.24B-parameter text model from mistralai on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.462score
Mistral-7B-Instruct-v0.2
mistralai
Open-weights 7.24B-parameter text model from mistralai on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.462score
Mistral-7B-v0.1
mistralai
Open-weights 7.24B-parameter text model from mistralai on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.397score
SmolLM-135M-Instruct-FP32
modularai
Open-weights 0.13B-parameter text model from modularai on Hugging Face.
ModelFFreshness: FOpen weights
0.367score
Kimi-Dev-72B
moonshotai
Open-weights 72.71B-parameter text model from moonshotai on Hugging Face (mit).
ModelFFreshness: FOpen weights
0.413score
Kimi-K2-Base
moonshotai
Open-weights 1026.41B-parameter text model from moonshotai on Hugging Face (modified-mit).
ModelDFreshness: DOpen weights
0.508score
Kimi-Linear-48B-A3B-Instruct
moonshotai
Open-weights 49.12B-parameter text model from moonshotai on Hugging Face (mit).
ModelDFreshness: DOpen weights
0.641score
Nanbeige4.2-3B
Nanbeige
Open-weights 4.17B-parameter text model from Nanbeige on Hugging Face (apache-2.0).
ModelBFreshness: BOpen weights
0.392score
DeepHermes-3-Llama-3-8B-Preview
NousResearch
Open-weights 8.03B-parameter text model from NousResearch on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.392score
Genstruct-7B
NousResearch
Open-weights 7.24B-parameter text model from NousResearch on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.392score
Hermes-2-Pro-Llama-3-8B
NousResearch
Open-weights 8.03B-parameter text model from NousResearch on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.392score
Hermes-2-Pro-Mistral-7B
NousResearch
Open-weights 7.24B-parameter text model from NousResearch on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights
0.462score
Hermes-3-Llama-3.1-8B
NousResearch
Open-weights 8.03B-parameter text model from NousResearch on Hugging Face (llama3).
ModelFFreshness: FOpen weights
0.432score
Hermes-4.3-36B
NousResearch
Open-weights 36.15B-parameter text model from NousResearch on Hugging Face (apache-2.0).
ModelDFreshness: DOpen weights
0.398score
Meta-Llama-3.1-8B
NousResearch
Open-weights 8.03B-parameter text model from NousResearch on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.462score
Meta-Llama-3.1-8B-Instruct
NousResearch
Open-weights 8.03B-parameter text model from NousResearch on Hugging Face (llama3.1).
ModelFFreshness: FOpen weights
0.367score
Nous-Hermes-13b
NousResearch
Open-weights 13B-parameter text model from NousResearch on Hugging Face (gpl).
ModelFFreshness: FOpen weights
0.398score
Nous-Hermes-2-Mixtral-8x7B-DPO
NousResearch
Open-weights 46.7B-parameter text model from NousResearch on Hugging Face (apache-2.0).
ModelFFreshness: FOpen weights

The weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.