gemma-4-26B-A4B-it-assistant-GGUF
ManniX-ITA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 ↗DISCOVER & COMPARE
Explore AI models 19
gemma-4-A4B-98e-v6-coder-it-GGUF
ManniX-ITA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 · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗gemma-4-A4B-98e-v7-coder-it-GGUF
ManniX-ITA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 ↗gemma-4-A4B-98e-v7-coderx-it-GGUF
ManniX-ITA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 · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗gemma-4-E2B-it-assistant-GGUF
ManniX-ITA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 ↗JackOD-9B-Coder
ManniX-ITA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 ↗JackOD-9B-Coder-MTP-DFlash-GGUF
ManniX-ITA**`ManniX-ITA/JackOD-9B-Coder` v3** quantized for **DFlash speculative decoding**: every quant tier of `JackOD-9B-Coder-MTP-GGUF` with the built-in MTP (NextN) layer removed, plus a separate **DFlash drafter** that does the drafting instead. Use **this** repo if your runtime drafts with DFlash. Use `JackOD-9B-Coder-MTP-GGUF` if it drafts with the model's own MTP head, or if you do not use speculative decoding at all. The two repos hold the same quantized weights; only the speculative-decoding path differs. JackOD-9B-Coder is a merge over `Qwen/Qwen3.5-9B` built for multi-turn agentic coding — DeltaCoder-9B 0.55 / Ornith-1.5-9B 0.30 / Coder-Flash 0.15, `omnimerge_v2`. The recipe, the evaluations and the full description are on the model card.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗JackOD-9B-Coder-MTP-GGUF
ManniX-ITAGGUF quants of **`ManniX-ITA/JackOD-9B-Coder`**, a four-way `omnimerge_v2` merge over `Qwen/Qwen3.5-9B` built for **multi-turn agentic coding** — staying on a task across turns and tool calls until it is actually finished, and then stopping. > **v3 (2026-10-24).** v3 is the v2 recipe with the two smaller weights swapped: > **DeltaCoder 0.55 / Ornith-1.5-9B 0.30 / Coder-Flash 0.15** (v2: Coder-Flash 0.30 / Ornith 0.15). > `omnimerge_v2` density 0.53, darex-q 0.75, seed 42, the root `eos_token_id` 248046 / `pad_token_id` 248055 > fix and the Coder-Flash tokenizer and template are unchanged. v3 ships a **new serving sampler** and was > evaluated with it; the v2 and v1 results kept at the end of this card were measured **greedy**, so they are > a record, not a same-basis comparison. Every `_K` and `IQ` tier is **imatrix-calibrated**, including `Q6_K`. The calibration matrix used to build them is published in this repo as `imatrix.dat` (sha256 `feb286c4…`) so the quants can be audited and reproduced.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗opencoti-llamafile
ManniX-ITASelf-contained single-file inference engine from the opencoti project — a Mozilla-Ocho llamafile base carrying the opencoti patch series: advanced KV residency/quantization (PolyKV, KVarN — see `docs/features/kvarn.md`), the KV rolling window (a KV cache larger than VRAM), elastic multi-session serving (elastic slots, an enforced KV admission gate with fast 429 + `Retry-After`, PolyKV shared-prefix pools with a REST control plane), DCA long-context extension, MTP speculative decode, sparse attention, RYS layer duplication, CUDA **and Vulkan** GPU backends, and a runtime introspection/control API. **Start with USAGE.md** — it explains exactly how this engine diverges from upstream llamafile, every added feature, its flags, defaults, limitations, and which features compose. This repo hosts the packaged release artifacts (they exceed GitHub's 2 GiB release-asset cap) plus the full patch series under `patches/` and the project documentation under `docs/`. Release notes and per-release `SHA256SUMS` live under `releases/` and are mirrored on the corresponding GitHub release at mann1x/opencoti.
Input Not verified→ Output Not verified
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Ornith-1.5-27B-A3B-Coder-MTP-GGUF
ManniX-ITAGGUF quantizations of **`ManniX-ITA/Ornith-1.5-27B-A3B-Coder`** — the pure competence-map 184-expert cut of `Ornith-1.5-35B-A3B`. **→ Recipe, drop maps, the full 11-bench evaluation, the cross-family comparison and the length/deliberation analysis all live on the weights card.** This card covers the quantized files only. > Its sibling **CoderX** > uses the same map with a REAP stability floor and measures better on almost every code and > reasoning axis at the same expert budget and the same file size. Start there unless you > specifically want this arm.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Ornith-1.5-27B-A3B-CoderX-MTP-GGUF
ManniX-ITA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 ↗Qwen3.6-27B-A3B-Coder
ManniX-ITA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 ↗Qwen3.6-27B-A3B-Coder-MTP-GGUF
ManniX-ITAA **code-specialist expert prune** of Qwen3.6-35B-A3B: the MoE is reduced from **256 experts to 184** (72 dropped per layer, ~35B→27B, still A3B active) using a **code-targeted competence map** (LiveCodeBench + MultiPL-E competence classes). Same router, attention, norms, MTP head and vision tower as the base — only the expert keep-set changes. **Served at top-10** (`num_experts_per_tok = 10`, baked as the default). This is a routing-recovery lever: after pruning to 184 experts, activating the top-10 (vs the base top-8) recovers instruction-following at **no cost to code** (see below). No fine-tuning, no distillation — pure expert selection + a routing-width dial. 1. **Competence map**: the 256e teacher is profiled per-expert on a balanced corpus + targeted **LiveCodeBench** and **MultiPL-E** (Rust/Java/JS) PASS-response classes. 2. **Drop map**: `wmax` aggregation with the LCB + MPE classes up-weighted (1.5) → 72/256 experts dropped per layer, protecting the code-competent experts. 3. **Top-10 routing** (`num_experts_per_tok = 10`) baked into the config → the shipped default. Pass `--override-kv qwen35moe.expert_used_count=int:8` to any llama.cpp tool to A/B back to native top-8.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Qwen3.6-27B-A3B-CoderX-MTP-GGUF
ManniX-ITAGGUF quantizations of **CoderX**, a long-horizon code prune of Qwen3.6-35B-A3B: the MoE is reduced from **256 experts to 184** (72 dropped per layer, ~35B→27B, still A3B active). Same expert budget as the sibling Qwen3.6-27B-A3B-Coder, but a different **selection** — same competence map, different keep-set. **Served at top-8** — the base model's native routing, no override needed. No fine-tuning, no distillation. CoderX is **not** a strict upgrade over Coder. It is the long-horizon sibling: far stronger on multi-step competitive-programming work, a tie on short multi-language completion, and its non-code profile is not yet measured. Pick by workload.
Input Text→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Qwen3.6-27B-Omnimerge-v4-GGUF
ManniX-ITAGGUF quantizations of `ManniX-ITA/Qwen3.6-27B-Omnimerge-v4` — the **MLP-passthrough** variant that defends against the Qwen3.6 think-policy fragility we discovered. Source dtype is BF16; this repo provides the standard bartowski quant ladder (F16 → IQ2_XXS) for `llama.cpp`. > **Source model:** `ManniX-ITA/Qwen3.6-27B-Omnimerge-v4` (BF16 weights, model card with full benchmarks and methodology). > **NOT** a quant of clean Qwen/Qwen3.6-27B — these GGUFs contain the v4 merge. > > **MTP companion (2× decode speedup):** weight-identical GGUFs with the MTP head retained for `llama.cpp --spec-type draft-mtp` self-speculative decoding are at `ManniX-ITA/Qwen3.6-27B-Omnimerge-v4-MTP-GGUF`. Quality is statistically indistinguishable from this repo (HE 137/164 ↔ 137/164, GPQA 155/198 ↔ 154/198); aggregate decode is 2.0-2.3 × faster on a single 24 GB GPU. Use that repo for interactive / single-request workloads where latency matters. All quants made using imatrix with calibration data v5, the same calibration set bartowski uses for the Qwen3.6 base release — so quality fingerprints are directly comparable to bartowski's `Qwen_Qwen3.6-27B-GGUF` repo.
Input Text · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Qwen3.6-27B-Omnimerge-v4-MTP-GGUF
ManniX-ITA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 · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Qwen3.8-27B-Omnimerge-v6-MTP-GGUF
ManniX-ITAGGUF quantizations of `ManniX-ITA/Qwen3.8-27B-Omnimerge-v6` — a task-arithmetic merge of three Qwen3.6 fine-tunes onto the **Qwen3.8-27B** base — with the **MTP head retained** (`blk.64.*`, 866 tensors) and the F16 vision projector published alongside. 19 tiers. `imatrix.dat` is archived in this repo, so the imatrix tiers are reproducible and auditable. This is a measured policy, not an omission. The imatrix crossover is a property of the quant **grid coarseness**: at the Q4 band and above the calibration bias outweighs importance-weighting and imatrix is neutral-to-harmful; below it (Q3/Q2/IQ) imatrix is the difference between a good quant and a broken one. Measured once per model family on a Q2→Q6 HE+/MPE ladder. So Q4_K_*/Q5_K_*/Q6_K* are built **without** imatrix and everything at Q3 and below — plus the IQ tiers — **with** it. Q8_0 needs none.
Input Text · Image→ Output Text
Price not verifiedCheck multimodal and channel pricing separately
View details ↗Ternary-Bonsai-2-27B-CRACK-MTP-GGUF
ManniX-ITA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 ↗Ternary-Bonsai-2-27B-MTP-GGUF
ManniX-ITAPrismML's Ternary Bonsai 2 27B (a 27B dense model with end-to-end ternary weights, Qwen3.8-27B hybrid backbone, 262K context, Apache-2.0) **with a NextN / MTP draft head bundled in**, so the model self-speculates — no second file, no external drafter. The published PrismML GGUFs carry no NextN layer. These bundles add the upstream Qwen/Qwen3.8-27B MTP head (its `mtp.*` tensors, exported by upstream llama.cpp `convert_hf_to_gguf.py --mtp --outtype q8_0`) as one extra layer; every other key and tensor is the PrismML file, byte for byte. The head stays in the normal basis (it is not among the Hadamard-rotated tensors); its token-embedding lookup on the rotated table gets the inverse transform in the engine (PrismML's `518ad108`). The vision projector is **not** duplicated here: use PrismML's `Ternary-Bonsai-2-27B-mmproj-Q8_0.gguf` (600 MiB) from the base repo with `--mmproj`.
Input Text→ Output Text
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.