China's high-end five-axis race is not one race — it is three, run by three very different players. KEDE (科德数控, 688305) is the technology pure-play: a fully in-house CNC brain and ~85% component self-sufficiency, small but deep. Haitian Precision (海天精工, 601882) is the volume leader: ¥3.37B revenue, gantry centers at 25–35% of the high-end domestic market, and the most concrete overseas network. Gree (格力智能装备) is the production-cell entrant: it pairs its own GNC controller with a self-made spindle + linear-motor motion stack and sells a whole cell, not a box. For an overseas buyer, the three are not competitors to rank — they are three fit profiles.
Any honest "Chinese five-axis" map needs at least three archetypes, because the question "who makes Chinese five-axis?" has three different answers depending on what you value:
This is the synthesis issue the series has been building toward. Issue #1 profiled KEDE; Issue #2 / the hub covered Haitian; Issue #15 covered the controller brain (GSK) that Gree's GNC also lives in; Issue #18 covered the spindle/linear-motor muscle Gree makes in-house. Here they meet in one table.
| Attribute | Detail (audited 2025) |
|---|---|
| Positioning | Pure-play five-axis; the only Chinese maker that builds the "brain" itself |
| Revenue / profit | ¥552M (−8.86%) / net ¥88.6M (−31.79%) |
| Gross margin | 39.48% (high-end CNC segment 42.4%) |
| R&D | ¥195M = 35.4% of revenue (+20.2%) — among the highest in the sector |
| Mix | High-end 5-axis CNC ¥331M (60%); automation lines ¥185M (+272.7%, 33%) |
| Self-sufficiency | In-house GNC + ~85% key components (the deepest stack of the three) |
| Caveat | Small scale; 2025 profit dipped on softer demand + more automation-line (lower-GM) mix |
KEDE is what "China closed the high-end gap" looks like at the component level: a CNC system, five-axis RTCP math and key parts all made in-house. Its 35% R&D ratio is the signal — it is buying depth, not volume. The trade-off is scale and cyclicality: at ¥552M revenue it is a fraction of Haitian's size, and 2025 shows the demand swing hits small pure-plays hardest.
| Attribute | Detail |
|---|---|
| Positioning | Appliance-giant entrant; sells machine + robot + thermal + process, not a box |
| Brain | Self-developed GNC controller (in-house, domestic) |
| Muscle | Self-made spindle + linear motor; self-developed swing head at ~−50% cost |
| Flagship | GA-FMB3020D dual-five-axis gantry (automotive die-casting programs, 120 m/min) |
| Supply | ~75% external-supply ratio; integration, not full in-house |
| Overseas | Building service (e.g. Malaysia); defensible +25% efficiency red line for export deals |
| Caveat | CNC is a segment of Gree Electric (000651); no separately disclosed CNC financials — rated Low |
Gree's bet is different from both peers: not "deepest brain" (KEDE) and not "most machines shipped" (Haitian), but the whole cell. Pairing GNC with a self-made motion stack lets Gree quote a turnkey line — robot loading, thermal management, process know-how — which is what production-scale buyers actually buy. The ~−50% self-made swing-head/linear-motor cost and +25% efficiency red line are the defensibility story for overseas.
| Attribute | Detail (audited 2025) |
|---|---|
| Positioning | Domestic gantry machining-center champion; import-substitution, going global |
| Revenue / profit | ¥3.37B (+0.48%) / net ¥429M (−17.97%) |
| Gross margin | 25.70% (gantry 27.6%, VMC 20.5%, HMC 29.7%) |
| Gantry share | ¥1.895B = 56.3% of revenue; high-end gantry domestic share ~25–35% |
| Overseas | ¥558M (+50%), 16.6% of total; overseas GM ~39% vs ~23% domestic |
| Self-made parts | Spindle, swing head, five-axis head, truss tool mag — import dependence falling |
| Global footprint | Mexico, Vietnam, Germany, Serbia + Brazil & 4 trading subs; "overseas capacity + local service" |
Haitian is the proof that Chinese high-end machines ship in volume and travel. ¥3.37B revenue (6× KEDE), a gantry line at 25–35% of the domestic high-end market, and an overseas business growing 50% with a ~39% overseas margin — plus real local subsidiaries, not just exporters. For a buyer who wants a reference site and service nearby, Haitian's network is the most concrete of the three today.
| Dimension | KEDE (688305) | Gree (格力智能装备) | Haitian (601882) |
|---|---|---|---|
| Core identity | Five-axis tech pure-play | Full production-cell entrant | Gantry volume leader |
| 2025 revenue | ¥552M | Not disclosed (Low) | ¥3.37B |
| 2025 net profit | ¥88.6M | Not disclosed (Low) | ¥429M |
| Gross margin | 39.5% | n/a | 25.7% |
| Brain | In-house GNC (~85% self) | In-house GNC | Buys/partners (self-made parts growing) |
| Muscle | In-house key parts | Self-made spindle + linear motor | Self-made spindle/swing head |
| Five-axis depth | Deepest (pure-play) | Strong (dual-gantry GA-FMB3020D) | Growing (five-axis extension) |
| Export readiness | Smaller | Building (Malaysia service) | Most concrete (global subs) |
| Best fit | Deep five-axis tech buyers | Turnkey cell buyers | Volume gantry / local-service buyers |
This three-way table is the clearest view yet of where Gree's model lands. It does not out-tech KEDE (KEDE's in-house GNC + 85% self-sufficiency is deeper) and does not out-scale Haitian (¥3.37B vs undisclosed). What Gree uniquely adds is the integrated production cell: GNC + self-made motion stack + robot + thermal + process, sold as one line. Three reads:
At ¥552M, KEDE is small and its 2025 profit fell ~32% on soft demand and a lower-margin automation-line mix. Deep tech, but demand swings hit it hardest. Verify a reference site and lead time for your exact model.
2025 net profit fell ~18% on domestic price competition; its horizontal machining center dropped 30%. Gantry is the rock (56% of revenue, stable margin); the rest is more contested. The overseas ~39% margin is real but still ~16% of the mix.
Gree's CNC financials are not separately disclosed, so scale/profitability can't be verified from public data — rate Low. The integration claim (GNC + self-made motion stack + cell) is strong positioning; ask for a running reference line and the +25% efficiency evidence before treating it as proven for your job.
| Claim | Source | Confidence |
|---|---|---|
| KEDE 2025: rev ¥552M, net ¥88.6M, GM 39.5%, R&D ¥195M (35.4%) | Audited annual report (688305) | High |
| KEDE in-house GNC + ~85% self-sufficiency; automation lines +272% | Audited report + company positioning | High |
| Haitian 2025: rev ¥3.37B, net ¥429M, GM 25.7%, gantry 56% | Audited annual report (601882) | High |
| Haitian overseas ¥558M (+50%), ~16.6%, overseas GM ~39% | Audited report + broker notes | High |
| Gree GNC + self-made spindle/linear motor; GA-FMB3020D; +25% efficiency | Company positioning / prior reporting | Medium |
| Gree standalone CNC financials (revenue/profit) | Not separately disclosed (subsidiary of 000651) | Low |
Match the contender to the job. KEDE for the deepest five-axis tech and an in-house brain. Haitian for high-volume gantry / work-center value with the strongest local service today (Germany, Serbia, Mexico, Vietnam, Brazil). Gree for a turnkey production cell with in-house motion stack and a defensible efficiency case. In every case, verify the exact model, controller tier, accuracy class and a local reference site.
KEDE (688305) and Haitian (601882) are the two verifiable listed windows — one a deep-tech small-cap at a 35% R&D ratio, the other a scaled, globally-expanding gantry leader with a ~39% overseas margin. Gree's CNC is not separately investable; treat it as a strategic read on where the integrated-cell model is heading, and watch Gree Electric (000651) disclosures for any future CNC breakout.
The three are also a make-vs-buy map. KEDE shows the in-house-brain extreme; Haitian shows scaled gantry + growing self-made parts; Gree shows the full-cell integration bet. For Gree's own supplier strategy, KEDE is the brain benchmark and Haitian the gantry/volume and overseas-service benchmark — two concrete external references for the same race Gree is running.
KEDE (科德数控). It is the only one of the three built around five-axis from the start — a fully in-house high-end CNC system (GNC) plus ~85% key-component self-sufficiency, and five-axis machines are the core of its revenue. Gree also does five-axis (dual-gantry GA-FMB3020D) but inside a broader machine+robot+thermal+process business; Haitian's core is gantry machining centers with five-axis as a fast-growing extension.
Haitian Precision (海天精工) is clearly furthest: 2025 overseas revenue ¥558M (+50%), ~16.6% of total, with overseas gross margin ~39% vs ~23% domestic, and a real overseas footprint (Mexico, Vietnam, Germany, Serbia, Brazil + trading subsidiaries). Gree is building overseas service (e.g. Malaysia) and leads with a defensible +25% efficiency red line; KEDE's overseas is smaller. For a buyer who wants local support now, Haitian's network is the most concrete.
Because Gree is the entrant changing the game structurally: it pairs its own GNC controller with a self-made spindle and linear-motor motion stack and sells a whole production cell (machine + robot + thermal + process), not just a box. That integration model is the strategic contrast this series tracks. We rate Gree's standalone CNC financials Low confidence (subsidiary of Gree Electric, 000651, not broken out) and rely on product/positioning facts we can stand behind.
Match the contender to the job. KEDE for the deepest five-axis tech and in-house brain. Haitian for high-volume gantry/work-center value with the strongest overseas service today. Gree for a turnkey production cell with in-house motion stack and a defensible efficiency case. Then verify the exact model, controller tier, accuracy class and a local reference site — the brand name is a starting point, not a specification.