Qwen3.5-2B-Base-GGUF
aashish1904基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗DISCOVER & COMPARE
Explore AI models 84836
Qwen3-8B-GGUF
aashish1904基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Embedding
Price not verifiedCheck multimodal and channel pricing separately
View details ↗基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Image
Price not verifiedCheck multimodal and channel pricing separately
View details ↗PatentSBERTa_V2
AAUBS基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Embedding
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Atlas-Pro-7B-Preview-1M-GGUF
aayanmishra-ml基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Not verified→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Not verified→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗r17-m133
abacada基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Not verified→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗bigstral-12b-32k
abacusai> In JavaScript, functions are first-class values, which means you can assign them to variables, pass them as arguments to other functions, and even nest them inside other functions. This leads to concise and expressive code. > > Here's an example of a simple recursive function that calculates the factorial of a given number in JavaScript: > > > > Now, let's try to write an equivalent code snippet in ANSI-C with similar functionality: > > > > As you can see, the ANSI-C code becomes more verbose and has several disadvantages compared to JavaScript: > > 1. Need to define a maximum recursion depth using a constant or an array index, which can lead to hardcoded limits and potential stack overflow issues. > 2. The use of preprocessor directives (#define) for defining functions and macros can make the code less readable and harder to understand. > 3. The lack of closures in ANSI-C requires using static variables or passing arguments as arguments to nested functions, making the code more verbose and less concise. > yaml dtype: float16 merge_method: passthrough slices: - sources: - layer_range: [0, 8] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [4, 12] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [8, 16] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [12, 20] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [16, 24] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [20, 28] model: mistralai/Mistral-7B-Instruct-v0.2 - sources: - layer_range: [24, 32] model: mistralai/Mistral-7B-Instruct-v0.2
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗bigyi-15b
abacusaiI recently made bigstral-12b and then I saw this new awesome model yi-9b and decided to make an embiggened version. Here is a recipe for Mai Tai:\n\n1: > 3 parts rum, 2: 3 parts pineapple juice, 3: half a cup of lime juice, 4: 6 to 8 fresh or frozen pineapple chunks, 5: crushed ice. > Mix all ingredients except ice and pour into glasses with ice. Garnish with a pineapple slice.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Llama-3-Smaug-8B
abacusaiThis model was built using the Smaug recipe for improving performance on real world multi-turn conversations applied to meta-llama/Meta-Llama-3-8B-Instruct. This version of Smaug uses new techniques and new data compared to Smaug-72B, and more information will be released later on. For now, see the previous Smaug paper: https://arxiv.org/abs/2402.13228.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Slerp-CM-mist-dpo
abacusaiThe model did achieve an improvement in TruthfulQA over `cookinai/CatMacaroni-Slerp` and GSM8K over `mncai/mistral-7b-dpo-v5` which was the goal of the merge leading to an average score that was a better than both. It is unclear why the TruthfulQA metric is still meaningfully lower than the base `mncai/mistral-7b-dpo-v5`.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-34B-v0.1
abacusaiThis model is a finetune of jondurbin's excellent bagel model. This model has not utilised any form of merging. We created Smaug-34B-v0.1 using a new fine-tuning technique, DPO-Positive (DPOP), and new pairwise preference versions of ARC, HellaSwag, and MetaMath (as well as other existing datasets). We introduce the technique and the full training details in our new paper: https://arxiv.org/abs/2402.13228. We show that on datasets in which the edit distance between pairs of completions is low (such as in math-based datasets), standard DPO loss can lead to a reduction of the model's likelihood of the preferred examples, as long as the relative probability between the preferred and dispreferred classes increases. Using these insights, we design DPOP, a new loss function and training procedure which avoids this failure mode. Surprisingly, we also find that DPOP outperforms DPO across a wide variety of datasets and downstream tasks, including datasets with high edit distances between completions.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-72B-v0.1
abacusaiWe recently released Smaug-72B-v0.1 which has taken first place on the Open LLM Leaderboard by HuggingFace. It is the first open-source model to surpass an average score of 80%. Smaug-72B is finetuned directly from moreh/MoMo-72B-lora-1.8.7-DPO and is ultimately based on Qwen-72B. We created Smaug-72B-v0.1 using a new fine-tuning technique, DPO-Positive (DPOP), and new pairwise preference versions of ARC, HellaSwag, and MetaMath (as well as other existing datasets). We introduce the technique and the full training details in our new paper: https://arxiv.org/abs/2402.13228.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-AI-Image-Detector
abacusaiSmaug-AI-Image-Detector sets a new state of the art for open-source AI-generated image detection. It is a fine-tune of Google's EmbeddingGemma-2 by Abacus.AI and the strongest open-source detector on real-world images. **A dedicated vision model, 41% smaller than its base.** We streamlined EmbeddingGemma-2 into a vision-only architecture by removing its 305M-parameter audio tower: 439M parameters instead of 744M, with no loss in accuracy. Architecture EmbeddingGemma-2 image encoder + linear classifier Total Parameters 439M (base model: 744M) Audio Tower Removed: vision-only architecture, −305M parameters Input Any image; scored as three 528 × 528 crops at native resolution Output Probability that the image is AI-generated Hardware NVIDIA GPU (~1 GB), Apple Silicon (Metal), or CPU Precision bfloat16 weights Formats PyTorch / sentence-transformers (this repo); GGUF for llama.cpp: Smaug-AI-Image-Detector-GGUF Base Model google/embeddinggemma-2 Adaptation Full fine-tune
Input Image→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-AI-Image-Detector-GGUF
abacusai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Image→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Flash
abacusai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Humanizer
abacusai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Humanizer-GGUF
abacusai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Llama-3-70B-Instruct
abacusaiThis model was built using a new Smaug recipe for improving performance on real world multi-turn conversations applied to meta-llama/Meta-Llama-3-70B-Instruct. The model outperforms Llama-3-70B-Instruct substantially, and is on par with GPT-4-Turbo, on MT-Bench (see below). EDIT: Smaug-Llama-3-70B-Instruct is the top open source model on Arena-Hard currently! It is also nearly on par with Claude Opus - see below.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Llama-3-70B-Instruct-32K
abacusaiThis is a 32K version of Smaug-Llama-3-70B-Instruct. It uses PoSE (https://arxiv.org/abs/2309.10400) and LoRA (https://arxiv.org/abs/2106.09685) adapter transfer. More details are coming soon. Score vs selected others (sourced from: (https://lmsys.org/blog/2024-04-19-arena-hard/#full-leaderboard-with-gpt-4-turbo-as-judge)). GPT-4o and Gemini-1.5-pro-latest were missing from the original blob post, and we produced those numbers from a local run using the same methodology. Note that we believe the number of tokens/verbosity of the model strongly influences the GPT-4 judge in this case, and at least partially explains the improvement in Arena-Hard score for the 32K model.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Mini
abacusai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Basic profile indexed · More details to follow
Input Text · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Smaug-Mixtral-v0.1
abacusaiThis model is part of the Smaug series of finetuned models. This one based on https://huggingface.co/mistralai/Mixtral-8x7B-v0.1 We use a new fine-tuning technique, DPO-Positive (DPOP), and new pairwise preference versions of ARC, HellaSwag, and MetaMath (as well as other existing datasets). We introduce the technique and the full training details in our new paper: https://arxiv.org/abs/2402.13228. We show that on datasets in which the edit distance between pairs of completions is low (such as in math-based datasets), standard DPO loss can lead to a reduction of the model's likelihood of the preferred examples, as long as the relative probability between the preferred and dispreferred classes increases. Using these insights, we design DPOP, a new loss function and training procedure which avoids this failure mode. Surprisingly, we also find that DPOP outperforms DPO across a wide variety of datasets and downstream tasks, including datasets with high edit distances between completions.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Minimax_h3_hybrid
abakanai基础档案已收录，详细介绍后续补充。可先查看厂商、分类与来源。
Input Image→ Output Video
Price not verifiedCheck multimodal and channel pricing separately
View details ↗No matching models
Adjust provider, category or input/output filters. New models are added continuously.
COST EXPLORER
Great models. Clear costs.
Adjust monthly token usage to compare estimated API costs.
1 M = 1 million tokens · USD
Compares base text-token rates within the same channel
Compares base text-token rates within the same channel
Uses collected, comparable USD-per-million-token rates. Snapshots exclude cache discounts, tiers, extra reasoning output and taxes. Check current source prices.
MODEL HANDBOOK
Understand before you choose
How should API prices be compared?+
Separate input, output and cache rates, then estimate the same workload. Image, audio and video rates may use items, duration or resolution rather than tokens.
Do open weights mean free commercial use?+
Not necessarily. Check each version’s license, commercial restrictions and policies. Self-hosting also incurs compute and maintenance costs.
Is the top-ranked model the best for me?+
Rankings depend on tasks, versions and methods. Compare accuracy, latency, reliability and costs on your own workload.