Roadmap

Roadmap#

FuriosaAI regularly publishes its software with new features, performance improvements, and expanded hardware support. This page shows the forward-looking roadmap of ongoing & upcoming projects and when they are expected to land, broken down by areas on our software stack.

Note

The latest release is 2026.1.0. You can find the release notes here.

Upcoming Releases 2026 Q1#

  • 🔨 Qwen3 MoE, GPT-OSS, K-EXAONE model support

  • 🔨 Qwen3 VL and multi-modal model support

  • 🔨 KV cache offloading support

  • 🔨 OpenResponse support

  • 🔨 Speculative decoding support

  • 🔨 PyTorch eager mode support

2025 Q3 - Q4#

Furiosa-LLM#

  • ✅ Hybrid batching support (i.e., chunked prefill or inflight-batching)

  • ✅ Exaone4, Qwen3 support

  • ✅ Guided-decoding support (libguidance, xgrammar backends)

  • ✅ Tool-calling support

  • ✅ Prefix-caching support

  • ✅ Pooling Model support (embedding, score, and rank)

  • ✅ Fine-tuned model support

  • ✅ Tensor Parallelism support Phase 2: Inter-chip

  • ✅ Hugging Face Hub support

  • ✅ Pre-compiled artifacts on Hugging Face Hub

  • ✅ Qwen2 and Qwen2.5 model support

  • ✅ EXAONE3 model support

  • ✅ API Key based authentication support

  • ✅ Harmony response format support

Quantization#

  • ✅ Fine-grained FP8 Quantization (dynamic quantization, mixed quantization)

Distributed & Scalable Inference#

  • ✅ llm-d integration

  • ✅ NPU operator support for Kubernetes

  • ✅ DRA (Dynamic Resource Allocation) support for Kubernetes

2025 Q1 - Q2#

  • ✅ Tool-calling support in Furiosa-LLM (2025.1.0 release)

  • ✅ Device remapping support (e.g., /dev/rngd/npu2pe0-3 -> /dev/rngd/npu0pe0-3) for container (2025.1.0 release)

  • ✅ Automatic configuration for the maximum KV-cache memory allocation (2025.1.0 release)

  • ✅ Min-p sampling support (2025.1.0 release)

  • ✅ Chunked Prefill support in Furiosa-LLM (planned for 2025.2.0 release)

  • ✅ Chat API support in Furiosa-LLM (planned for 2025.2.0 release)

  • ✅ Reasoning parser support (2025.2.0 release)

  • ✅ Torch 2.5.1 support (2025.2.0 release)

  • ✅ Python 3.11 and 3.12 support (2025.2.0 release)

  • ✅ Support for building bfloat16, float16, and float32 models to model artifact without quantization (2025.2.0 release)

  • ✅ Metrics endpoint (/metrics/) support in Furiosa-LLM (2025.2.0 release)

  • ✅ Model artifact support in Huggingface Hub (2025.2.0 release)

  • ✅ Sampling parameter “logprobs” support (2025.2.0 release)

  • ✅ Container Runtime and Container Interface Device (CDI) support (2025.2.0 release)

2024 Q4#

  • ✅ Language Model Support: CodeLLaMA2, Vicuna, Solar, EXAONE-3.0 (2024.2.0 release)

  • ✅ Vision Model Support: MobileNetV1, MobileNetV2, ResNet152, ResNet50, EfficientNet, YOLOv8m, etc (2024.2.0 release)

  • ✅ Tensor Parallelism support Phase 1: Intra-chip (2024.2.0 release)

  • ✅ Torch 2.4.1 support (2024.2.0)

  • ✅ Huggingface Optimum integration (2024.2.0 release)