OpenModelsHQ

Text models 314

We track 314 open text models from 35 labs, 87 of them released in the 12 months to 7 Nov 2026. The newest is Qwen3.8-27B-DSpine-b16 (7 Nov 2026) and the most downloaded is Qwen3-0.6B, with 19,141,315 downloads in 30 days. 172 fit a 16 GB GPU at 4-bit, and 43 have a hosted API.

llama.cpp, vLLM and MLX provide different ways to run text-generation models. Local use favours easy installation and memory fit; a shared endpoint also needs to handle concurrent requests. The model architecture, quantisation and runtime must be compatible before a published benchmark is useful.

What the category covers

Instruction-following chat models are the obvious case, but the same architecture class includes base models meant for further training, reasoning variants tuned to spend more tokens before answering, and small models in the sub-billion range aimed at phones and edge boxes. Mixture-of-experts designs sit in the same category: they change which parameters activate per token, not what the model consumes or emits.

How they are measured

EleutherAI’s Language Model Evaluation Harness provides a shared framework for evaluating generative language models across academic tasks. Its repository documents task definitions, model integrations and configuration controls. A common tool helps researchers reproduce an evaluation, but it does not make results from different settings interchangeable.

That matters more than any individual score. Benchmark numbers in a model card are only comparable when the harness, the few-shot count and the prompt template match, and the harness exists precisely so those conditions can be pinned down. When a lab reports a number without naming the evaluation setup, treat it as a claim rather than a measurement.

Scores also decay in usefulness as they leak into training data. Contamination-aware benchmarks address this by collecting fresh problems over time rather than publishing a fixed set, an approach LiveCodeBench applies on the coding side.

What serves them

vLLM supports GPU serving through CUDA, ROCm and Intel XPU paths, with Apple GPU support through the separate vLLM-Metal plugin. Its installation overview also lists CPU targets and other hardware plugins. Each backend has its own installation and model-support limits.

llama.cpp takes the other approach, targeting local and cloud inference across a range of hardware. Its feature list includes integer quantisation and CPU plus GPU hybrid inference to partially accelerate models larger than total video memory. Getting a model running is one command: llama cli -hf with a repository name, or llama serve -hf for an OpenAI-compatible server.

On Apple silicon, mlx-lm is the native path, described by its README as a Python package for generating text and fine-tuning language models, with Hugging Face Hub integration, quantisation and LoRA. Our MLX page covers the commands.

Ollama wraps llama.cpp-style local running in a simpler interface, covered on our Ollama page, and the GGUF format those tools consume is explained on our GGUF page.

The hardware reality

Weights have to fit somewhere, and memory limits you before compute does. A model that does not fit in video memory either spills to system memory, which llama.cpp supports through hybrid inference at a speed cost, or does not run. Quantisation is the usual lever, trading some accuracy for a smaller footprint.

Apple silicon behaves differently because of unified memory: MLX arrays live in shared memory, so a Mac with more RAM can load larger models without a separate video memory budget. Memory bandwidth and runtime overhead still affect speed.

For serving rather than single-user chat, memory also goes to the key-value cache. vLLM’s optimization guide says cache space can run short when requests are batched, in which case it preempts and recomputes requests, and its memory guide lists a lower context length and a smaller maximum batch size among the ways to reduce memory use. A model that runs as a local chat assistant can need different settings as a shared endpoint on the same card.

Licence patterns

Model licences differ. For the conditions attached to permissive releases, see our pages on Apache 2.0, MIT and the comparison. Custom community licences add conditions such as naming requirements or use restrictions, which is the pattern on our Llama Community Licence and Gemma terms pages. A third group uses MIT with additions, covered on our modified MIT page.

Check the licence file in the repository rather than the label on the model card, because the two disagree often enough to matter.

This summarises licence text. It is not legal advice.

FAQ

Which runtime should I start with?

For a local machine, llama.cpp or Ollama, because quantised weights and hybrid inference make modest hardware workable. For a GPU server with concurrent users, vLLM. On a Mac, MLX.

Are benchmark scores in model cards comparable?

Only when the evaluation setup matches. The Language Model Evaluation Harness exists to make runs reproducible, and cards that do not state the harness, task version and shot count are not comparable to each other.

Do mixture-of-experts models need less memory?

Fewer active parameters can reduce computation per token, but that does not reduce the full checkpoint to its active-parameter size. Memory use also depends on quantisation, expert placement and offloading.

Can one model handle text and images?

Some do, and those are a separate category on this site. See our vision-language page for how they are measured and served.

Pick a view

All models

Newest

ModelReleasedTypeSizeMemory at 4-bit16 GB GPUDownloads, 30 daysCheapest API, blended / 1M
7 Nov 2026LLM2B1.5 GBFits4,332No tracked API
26 Oct 2026LLM11.2B8 GBFits189No tracked API
21 Oct 2026LLM286B189.7 GBToo big3,372No tracked API
21 Oct 2026LLM286B189.7 GBToo big1,157No tracked API
17 Oct 2026LLM7.8B5.5 GBFits170No tracked API
15 Oct 2026LLM30.1B20.7 GBToo big7,755No tracked API
14 Oct 2026LLM8.8B7.3 GBFits6,349No tracked API
14 Oct 2026LLM30B22.2 GBToo big5,081No tracked API
2 Oct 2026LLM78.1B51.7 GBToo big2,894No tracked API
27 Sep 2026LLM311B206.4 GBToo big20,694No tracked API
27 Sep 2026LLM1.02T679.9 GBToo big20,162No tracked API
21 Sep 2026LLM311B206.4 GBToo big57,506$0.14 (₹13)

Most downloaded

ModelReleasedTypeSizeMemory at 4-bit16 GB GPUDownloads, 30 daysCheapest API, blended / 1M
27 Apr 2025LLM752M1.5 GBFits19,141,315No tracked API
27 Apr 2025LLM14.8B11.2 GBFits13,011,347$0.13 (₹12)
1 Dec 2025LLM685B468.1 GBToo big8,442,070$0.234 (₹22)
16 Sep 2024LLM7.6B5.5 GBFits6,444,573No tracked API
27 Apr 2025LLM8.2B6.7 GBFits6,165,334No tracked API
4 Aug 2025LLM20.9B14 GBFits5,673,703$0.0225 (₹2.12)
17 Sep 2024LLM1.5B1.3 GBFits5,145,856No tracked API
16 Sep 2024LLM494M0.4 GBFits4,292,272No tracked API
18 Jul 2024LLM8B5.7 GBFits4,198,546No tracked API
27 Apr 2025LLM32.8B24 GBToo big3,741,049No tracked API
10 Mar 2025LLM1000M0.7 GBFits3,406,623No tracked API
4 Aug 2025LLM117B77.4 GBToo big2,433,479$0.045 (₹4.24)

Smallest that fit 16 GB

ModelReleasedTypeSizeMemory at 4-bit16 GB GPUDownloads, 30 daysCheapest API, blended / 1M
2 Dec 2025LLM0M1.2 GBFits543No tracked API
13 Oct 2024LLM9M0 GBFits5,899No tracked API
13 Oct 2025LLM32M0 GBFits255No tracked API
17 Apr 2024LLM128M0.1 GBFits8,685No tracked API
13 Oct 2025LLM186M0.1 GBFits152No tracked API
7 Feb 2024LLM248M0.3 GBFits1,308No tracked API
15 Apr 2024LLM248M0.3 GBFits200No tracked API
7 Feb 2024LLM248M0.3 GBFits136No tracked API
11 Apr 2024LLM248M0.3 GBFits129No tracked API
28 Jun 2025LLM361M0.4 GBFits8,875No tracked API
28 Jun 2025LLM361M0.4 GBFits2,113No tracked API
30 Mar 2026LLM438M0.5 GBFits2,840No tracked API
Top 12 in each view. Memory is the 4-bit estimate with an 8k context; a model fits when it stays within 90% of the GPU's memory.How we calculate

Releases per month

87 in 12 months
Text model releases per month
Months run to the newest release in our data, across all model types. Small screens show the last 6 months; the data table has all 12.
Text model releases per month: data
MonthTotal
Dec 202511
Jan 20266
Feb 20269
Mar 20267
Apr 202610
May 20264
Jun 202610
Jul 20265
Aug 202610
Sep 20266
Oct 20268
Nov 20261

Cheapest hosted APIs

All API prices
  1. gpt-oss-20bOpenAI · 20.9B · PrimeIntellect$0.0225 (₹2.12)
  2. DeepSeek-V4-FlashDeepSeek · 291B · Baidu$0.0314 (₹2.96)
  3. gpt-oss-120bOpenAI · 117B · PrimeIntellect$0.045 (₹4.24)
  4. NVIDIA-Nemotron-3.5-Lightning-30B-A3B-BF16NVIDIA · 31.6B · Darkbloom$0.0488 (₹4.60)
  5. c4ai-command-r7b-12-2024Cohere Labs · 8B · Cohere$0.0656 (₹6.19)
  6. phi-4Microsoft · 14.7B · DeepInfra$0.07 (₹6.60)
  7. Llama-3.2-1B-InstructMeta · 1.2B · Cloudflare$0.0705 (₹6.65)
  8. granite-4.2-8bIBM Granite · 8.8B · CoreWeave$0.0825 (₹7.78)

Blended is 3 parts input to 1 part output, USD per million tokens, with INR at ₹94.29 per $1.

Top labs for text

Most tracked text models first

Sources

Other model types